Analysis of Performance Differences Between Self-Developed Fundus Surgery Image Evaluation Software and Kinovea in Ophthalmic Microsurgery

Analysis of Performance Differences Between Self-Developed Fundus Surgery Image Evaluation Software and Kinovea in Ophthalmic Microsurgery | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Analysis of Performance Differences Between Self-Developed Fundus Surgery Image Evaluation Software and Kinovea in Ophthalmic Microsurgery Haonan Xu, Yunwei Hu, Yincong Xu, Pisong Yan, Bowen Zhong, Weifeng Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8029636/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Objective This study systematically compares the performance of the self-developed "Fundus Surgery Image Evaluation Software" and the open-source tool Kinovea in quantifying key parameters of ophthalmic microsurgical operations. Methods The self-developed "Fundus Surgery Image Evaluation Software" integrates modules for video cropping, key point annotation, reference frame verification, and automatic output, enabling simultaneous measurement of needle insertion depth and tremor amplitude. The performance of this software was evaluated through validation experiments (in vitro models, ex vivo porcine eyes, and clinical surgical videos) and compared with that of Kinovea. For the in vitro models, standard needle insertion depths of 300–700 µm and standard tremor values of 50–250 µm were set; ex vivo porcine eyes were used to simulate an environment similar to the human retinal environment, with measurements of needle insertion depth and tremor values within the same ranges; and clinical surgical videos were analyzed based on the dynamic retinal background in living subjects. The main measurement indicators included needle insertion depth, as well as the mean, maximum, minimum, median, and variance of tremor. The measurement deviation and stability of the two software tools were compared. Results In all experimental scenarios, the measurement results of the self-developed "Fundus Surgery Image Evaluation Software" were significantly superior to those of Kinovea. In the in vitro scenario, the needle insertion error of the self-developed software was < 0.5%, while Kinovea exhibited a systematic positive bias of 40%–68%. For the tremor segment, the self-developed software showed a slight underestimation but could identify tremors throughout the entire process, whereas Kinovea barely detected tremors (mean value < 5 µm). In the ex vivo porcine eye experiment, the needle insertion deviation of the self-developed software was ≤ ± 20%, the mean tremor value was < 160 µm, and stability was maintained across segments. In contrast, Kinovea consistently overestimated the needle insertion depth and produced an extreme value of 1,500 µm; its tremor measurement was interfered with by background textures, leading to a 2–8-fold overestimation and even an abnormal spike of 35,618 µm. In the clinical video analysis, the needle insertion depth measured by the self-developed software ranged from 198 to 934 µm, with a mean tremor value < 250 µm and a maximum tremor value < 800 µm, all falling within the safe range. For Kinovea, the maximum needle insertion depth reached 2,107 µm and the maximum tremor value was 2,101 µm; 8 out of 10 cases showed a systematic overestimation of 1.2–4.6 folds. Conclusion The self-developed "Fundus Surgery Image Evaluation Software" maintains high accuracy, low dispersion, and values within the clinically acceptable range under blank, tissue-based, and in vivo complex background conditions, making it safe for application in the quantitative evaluation of retinal microsurgery. Due to limitations in template matching, Kinovea exhibits significant systematic biases and extreme abnormalities, and thus is not suitable for direct use in the measurement of ophthalmic microsurgical procedures. Fundus Surgery Image Evaluation Software Kinovea Needle Insertion Depth Tremor Value Ophthalmic Microsurgical Manipulation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Introduction Subretinal injection refers to the direct delivery of drugs into the potential space between the neurosensory retina and the retinal pigment epithelium (RPE) layer, namely the subretinal space. It is an administration method that delivers therapeutic drugs to the subretinal space [ 1 ] . Currently, the most common administration route for fundus diseases remains intravitreal injection, where drugs are directly injected into the vitreous humor, enabling rapid achievement of high concentrations. However, drugs may be cleared relatively quickly from the vitreous humor, necessitating frequent administrations. Compared with intravitreal injection, subretinal injection—an administration method that has developed rapidly in recent years—has become increasingly prevalent in ophthalmic clinical practice and research. It offers advantages including direct drug action, high local drug concentration, and prolonged duration of interaction with the retina [ 2 ] , while minimizing the risks of immune rejection and inflammation [ 3 ] . Subretinal injection requires the surgeon to accurately penetrate the neurosensory retina (with a thickness of only 150–200 µm) using an injection needle without damaging the retinal pigment epithelium (RPE) layer [ 4 ] . Additionally, it is necessary to ensure a certain level of stability in needle insertion depth, injection volume, and injection pressure to prevent further retinal damage [ 2 , 5 , 6 ] . These requirements make the technique highly demanding on the surgeon’s operational skills, highlighting the critical importance of a dedicated device capable of maintaining excellent stability. Surgical robots can to a certain extent address the issue that it is difficult for surgeons to maintain a high standard of stability during subretinal injection. Combined with microsurgical technology, surgical robots can assist ophthalmologists in performing more precise operations during surgery, maintain operational stability, and simultaneously reduce secondary damage to normal tissues during the surgical process [ 7 , 8 ] . Currently, ophthalmic surgical robots that have been applied in clinical practice have been proven to exhibit less tremor compared with manual operations during retinal surgery; they can overcome the physiological limitation of surgeon fatigue and maintain the stability of surgical instruments [ 9 , 10 ] . However, we have found that there is a lack of quantitative analysis on the operational path of surgical robots, and there is also an absence of a relatively standard operational model to serve as a comparison benchmark. Against this background, for beginners in ophthalmic surgery, the traditional training model lacks a quantitative evaluation system, making it difficult to help them accurately master the key points of surgery. In the field of surgical robot research and development, a standardized operational model is also required to improve the accuracy of automated systems. With the in-depth application of artificial intelligence (AI) technology in the medical field, quantitative analysis of surgical videos has become a key approach to solving the aforementioned problems. On the one hand, by leveraging computer vision and AI technologies to extract spatiotemporal features and analyze motion sequences from expert surgical videos, complex surgical operations can be converted into quantifiable parameter indicators. This provides beginners with intuitive learning feedback and improvement directions, accelerating the process of skill acquisition. On the other hand, these quantitative analysis results can also provide surgical robots with references for standard operating methods and precise path planning, facilitating the construction of more intelligent and safer automated surgical assistance systems and promoting the advancement of subretinal injection technology toward standardization and intelligentization. Currently, we have observed that Kinovea software is used in rehabilitation medicine, sports medicine, biomechanical research, and clinical evaluation to analyze motion trajectories and force conditions. Kinovea is an open-source video analysis software (specializing in 2D motion tracking and biomechanical analysis), primarily applicable to basic video analysis and sports biomechanical analysis. It has been widely adopted due to its free and open-source nature, low learning threshold, and relatively professional biomechanical analysis capabilities. However, its limitations—including limited precision, low automation, lack of 3D analysis, and insufficient medical adaptation—indicate that it is unsuitable for high-precision ophthalmic surgeries. To date, no team has developed a software specifically tailored for ophthalmic surgical procedures. Therefore, a software capable of analyzing high-precision ophthalmic surgeries and capturing surgical movements for quantitative analysis is particularly crucial. Methods 1.1 Software System The self-developed Fundus Surgery Image Evaluation Software is a high-precision analysis tool specifically designed for ophthalmic surgeries. Its primary function is to quantify key parameters during ophthalmic surgical procedures, such as needle insertion depth measurement and tremor magnitude, thereby providing data support for analyzing the accuracy of surgical operations. The Fundus Surgery Image Evaluation Software consists of four core modules: 1. Video Selection: Including rough clipping, surgical video selection, playback, pause, and video segment clipping; 2. Pre-analysis: Encompassing data annotation and fine adjustment of marked points; 3. Reference Frame: Involving image selection and reference frame verification; 4. Data Analysis: Covering predicted point visualization, indicator data visualization, and report output. The basic operational workflow is as follows: 1. Rough Clipping: Extract key steps from the entire surgical video, including sequential operations such as pre-insertion, injection, fundus elevation, and post-withdrawal, while excluding invalid frames. Drag the surgical video from the folder into the editing software project file and add it to the timeline. Confirm the video segment by dragging the scale bar, click the "Clip" button to perform clipping, select and delete unwanted clipped segments, and finally click the "Export" button in the toolbar to select the output video format and save path. 2. Video Selection: Click "Select Video Source" to choose and open the surgical video. Then click to select the corresponding video type, and mark whether the current video is recorded from manual operation or robotic operation. 3. Video Frame Selection: Determine the standard positions of the left and right selection frames based on surgical movements. For example, in the "subretinal injection evaluation" mode, the left and right frame ranges can be defined using the keyframe method, by dragging the progress bar to record the left and right frame indices, or by manually entering the indices for frame selection. Click the "Pre-analysis" button after completing frame selection. 4. Pre-analysis: Annotate the microneedle point on the instrument detection frame. Click "Open Schematic" to view the position of the marked point, switch between images using the "Previous" and "Next" buttons, and click "Start Calibration" to calibrate the position. The image can be zoomed in or out using the shortcut "Ctrl + Mouse Wheel". Click "Start Pre-analysis" after calibration is completed; if there are redundant selection frames or missing marked points, corresponding adjustments can be made. 5. Reference Frame: Drag an image that meets the criteria of complete and clear needle tip, and close proximity to the retina, to set it as the reference frame. Click "Verify Reference Frame" to perform verification; the image can be zoomed using the mouse wheel. After confirming that the image meets the criteria, check the selection box and click "Verify". 6. Data Analysis: Click "Start Analysis" to initiate the analysis process, and click "Stop Analysis" to terminate it if needed. During the analysis, visualization of predicted points and evaluation indicators can be performed. Click the "Play" button to control the playback and pause of the data video, and drag the progress bar to control the display progress. Finally, click "Export Report", fill in the required information, and click "Print" or "Save" to complete the report output. Kinovea is a feature-rich and powerful open-source tool for video recording, measurement, and annotation. It supports image visualization, recording, measurement, annotation, and (high-speed) motion analysis, enabling users to capture, compare, annotate, and measure movements in videos. It has a wide range of application scenarios, including sports, laboratory settings, and industrial automation. The basic operational workflow is as follows: 1. Image or Video Acquisition Launch the Kinovea software, locate the camera name in the software interface and double-click it to obtain the real-time image from the camera. To import a video from a file, you can use File Explorer to navigate to the folder where the video is stored, then double-click the thumbnail of the video file in the thumbnail panel on the right to open the video in the player screen; alternatively, you can use the "File > Open" menu option, or directly drag the video file from Windows Explorer to the Kinovea software interface to open it. 2. Video Operations (1) Playback Control: After opening the video, use the control buttons of the player to perform operations such as starting playback, pausing, fast-forwarding, and rewinding; you can also drag the navigation cursor to jump to any position in the video. (2) Workspace Setup: Navigate to the start position of the segment you are interested in within the video and click the "Workspace Start" button; then play the video until the end of the segment requiring analysis, and click the "Workspace End" button. After setting up the workspace, if the loop playback mode is enabled, the video will play in a loop within the workspace, and the navigation cursor will be more precise within the workspace at this time. (3) Image Size Adjustment: If the video image is too small, you can drag the small squares at the corners of the image to adjust its size; alternatively, you can directly perform zoom operations on the image or use the magnifier tool to enlarge image details. (4) Playback Speed Adjustment: If a more detailed study of movements is needed, you can reduce the playback speed using the speed slider. To restore normal speed (100%), simply double-click the speed percentage value. You can also use the up and down arrow keys on the keyboard to adjust the speed in small increments. 3. Image and Video Annotation (1) Distance Measurement: Click the line button in the annotation tools and draw a line on the image. After drawing, left-click the line and select "Show Measurements > Length", at which point the pixel length of the line will be displayed. To obtain accurate measurements in actual millimeters, it is necessary to place an object of known size (e.g., a ruler) in the image. Draw a line corresponding to the known actual length, left-click the line, select "Calibrate", enter the actual length of the line, and click "Apply". All subsequent lines drawn will display their actual lengths based on this calibration. (2) Adding Annotations: Kinovea provides a variety of annotation tools, such as those for adding arrows, text descriptions, and drawing shapes, which are used to highlight key positions or features in the video. Simply click the corresponding annotation tool button and draw or add content directly on the video screen. (3) Adding Key Images: When the video plays to a key position, click the "Add Key Image" button to mark the current frame as a key image. Thumbnails of key images will be displayed in the software interface, facilitating subsequent review and comparison. 4. Data Export and Saving After completing video analysis and annotation, the analysis data can be exported to spreadsheet formats (e.g., OpenOffice Calc, Excel, or plain text files) for further processing and analysis. Through the corresponding export menu options, select the type of data to be exported (e.g., time, length, angle values) to perform the export. If it is necessary to save the video or its key images, select the "Save" option via the "File" menu, and choose the save format and path according to the prompts. For the saved video, if there are drawn and annotated contents during the analysis process, you can select whether to embed these contents into the video based on the software settings. 1.2 Validation Experiment Design This study systematically evaluated the performance of the self-developed Fundus Surgery Image Evaluation Software and the conventional Kinovea software in measuring ophthalmic surgery-related operational data through a three-stage data comparison. 1.3 Research Subjects and Data Sources Three types of data were selected as research samples: first, in vitro simulation data (including a needle insertion group and a tremor group, each divided into five subgroups. For the needle insertion group, standard needle insertion depths were set to 300 µm, 400 µm, 500 µm, 600 µm, and 700 µm; for the tremor group, standard tremor values were set to 50 µm, 100 µm, 150 µm, 200 µm, and 250 µm); second, ex vivo porcine eye experimental data (the needle insertion group was divided into five subgroups with dynamic tremor data, and standard needle insertion depths for the five subgroups were set to 300 µm, 400 µm, 500 µm, 600 µm, and 700 µm); third, ophthalmic surgical videos (containing dynamic processes related to needle insertion and tremor). 1.4 Software Operation and Index Measurement The aforementioned data were analyzed using the Fundus Surgery Image Evaluation Software and Kinovea software, respectively. The measured indices included needle insertion depth (µm), mean tremor value (µm), maximum tremor value (µm), minimum tremor value (µm), median tremor value (µm), and tremor variance (µm). All operations were performed by professional personnel to ensure the consistency of the measurement process. 1.5.1 In Vitro Model The in vitro model was used to measure the standard needle insertion depth and standard tremor value set for the instrument under a fixed blank background. The purpose of this model was to compare the accuracy of the measurement data from the two software programs in a simple background, excluding the interference of complex backgrounds and background movement. The in vitro data were divided into a needle insertion group and a tremor group, which were used to measure needle insertion and tremor, respectively; the tremor in the needle insertion group was only for reference and control. The needle insertion group was further divided into five subgroups, with standard needle insertion depths set to 300 µm, 400 µm, 500 µm, 600 µm, and 700 µm; each standard needle insertion depth was measured repeatedly for 5 times. The needle insertion depth was measured using the self-developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively. The tremor group was also divided into five subgroups, with standard tremor values set to 50 µm, 100 µm, 150 µm, 200 µm, and 250 µm; each standard tremor value was measured repeatedly for 5 times. The tremor values were measured using the self-developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively(Fig. 1 ). 1.5.2 Ex - vivo Porcine Eye The ex vivo porcine eyes used in this study were by-products of food processing from qualified abattoirs, with sources complying with national relevant regulations and animal welfare requirements. The ex - vivo porcine eye was used to measure the standard needle insertion depth of the instrument into the fixed retina within the porcine eye and the tremor during the operation process. The aim was to compare the accuracy of the measurement data of the two groups of software in a background relatively close to the human retina, but without the interference of respiration and eye movement. The needle insertion group was divided into five subgroups, with standard needle insertion depths set to 300 µm, 400 µm, 500 µm, 600 µm, 700 µm, 800 µm, and 900 µm. Each standard needle insertion depth was measured repeatedly for 5 times, and the corresponding tremor value during each needle insertion was measured simultaneously. The measurements were performed using the self - developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively(Fig. 2 ). 1.5.3 Clinical Surgical Videos Ten ophthalmic surgical videos were selected, and the depth of instrument needle insertion into the retina and tremor were measured against the background of the patient’s moving retina. Measurements were performed using the self-developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively. This group represents real clinical scenarios, with the living retina as the background—characterized by complex and moving images. The software was required to accurately identify the instrument against the complex background and eliminate interferences such as respiratory movement or eye tremor(Fig. 3 ). Results 2.1 Data Processing Results of the In Vitro Model The experimental results of the needle insertion group are as follows (Fig. 4 ): In the standard depth segment of 300 µm (Serial Numbers 1–5), the measured values of the self-developed Fundus Surgery Image Evaluation Software were concentrated in the range of 299.6–305.6 µm, closely aligning with the blue dashed line. In contrast, the measured values of Kinovea were significantly higher (411.8–495.6 µm), deviating from the standard line by approximately 40%–65%. In the 400 µm segment (Serial Numbers 6–10), the self-developed Fundus Surgery Image Evaluation Software continued to maintain measurements within the narrow range of 399.2–402.0 µm, almost overlapping with the standard value; however, the results of Kinovea further increased to 583.9–639.8 µm, with a deviation > 45%. In the 500 µm segment (Serial Numbers 11–15), the measured values of the self-developed Fundus Surgery Image Evaluation Software remained stable at 499.98–501.7 µm, with an error < 0.35%; in comparison, the results of Kinovea had increased to 734.6–830.1 µm, showing a deviation of approximately 47%–66%. In the 600 µm segment (Serial Numbers 16–20), the self-developed Fundus Surgery Image Evaluation Software yielded readings of 600.0–602.0 µm, maintaining consistency with the standard line; whereas Kinovea reached 903.5–998.5 µm, with a deviation > 50%. In the 700 µm segment (Serial Numbers 21–25), the self-developed Fundus Surgery Image Evaluation Software measured 698.8–703.6 µm, with an error < 0.5%; the results of Kinovea had risen to 1068.9–1173.4 µm, achieving a deviation of 50%–68%. In summary, under the simple in vitro background, the measured needle insertion depths obtained by the self-developed Fundus Surgery Image Evaluation Software almost completely matched the set standards, demonstrating high accuracy. In contrast, Kinovea software exhibited systematic positive deviations under all standard depth conditions, and these deviations continued to expand as the set depth increased. This finding indicates that Kinovea still has significant range-related errors even in the absence of background interference. Results of the Tremor Group (Fig. 5 ): In the standard 50 µm segment (Serial Numbers 1–5), the mean values measured by the self-developed Fundus Surgery Image Evaluation Software were stable at 46.9–48.1 µm, closely adhering to the standard line. In contrast, the mean values of Kinovea were only 1.63–3.05 µm, with a maximum value of merely 6.4–18.0 µm, which was significantly lower than the standard value—indicating that Kinovea could barely distinguish low-amplitude tremors. In the 100 µm segment (Serial Numbers 6–10), the self-developed Fundus Surgery Image Evaluation Software yielded mean values of 86.5–88.5 µm and maximum values of 96.3–101.9 µm, still remaining close to the standard. However, Kinovea’s mean values were 1.63–3.95 µm and maximum values were 5.56–15.8 µm, which were still far below the standard, demonstrating that Kinovea still lacked sufficient sensitivity in this segment. In the 150 µm segment (Serial Numbers 11–15), the self-developed Fundus Surgery Image Evaluation Software had mean values of 123.1–130.1 µm and maximum values of 139–144 µm, showing a slight underestimation but remaining identifiable. Kinovea, on the other hand, had mean values of 2.00–4.09 µm and maximum values of 9.76–17.1 µm, which were far below the standard and showed no significant difference from those in the 50 µm and 100 µm segments—further confirming that Kinovea could not effectively capture low-to-moderate amplitude tremors. In the 200 µm segment (Serial Numbers 16–20), the self-developed Fundus Surgery Image Evaluation Software produced mean values of 168.1–174.2 µm and maximum values of 179–190 µm, approaching the standard. Kinovea’s mean values were 2.11–3.09 µm and maximum values were 8.37–15.6 µm, still maintaining an extremely low level—suggesting that Kinovea’s ability to identify high-frequency signals was continuously limited. In the 250 µm segment (Serial Numbers 21–25), the self-developed Fundus Surgery Image Evaluation Software had mean values of 217.2–218.2 µm and maximum values of 230–236 µm, with a slight underestimation but a consistent trend. Kinovea’s mean values were 2.28–3.29 µm and maximum values were 6.85–14.7 µm, which were almost the same as those in the previous segments—indicating that Kinovea could not effectively detect tremors throughout the entire measurement range. In summary, under the simple in vitro background, the self-developed Fundus Surgery Image Evaluation Software could effectively identify standard tremors ranging from 50 to 250 µm, with only a slight underestimation observed in the maximum amplitude segment. In contrast, the mean and maximum tremor values measured by Kinovea software in all standard segments were significantly lower than the standard values and remained at an extremely low level of < 5 µm. This indicates that Kinovea has insufficient resolution for tremor signals and cannot detect the actual tremor amplitude. 2.2 Data Processing Results of Ex Vivo Porcine Eyes The measurement results of needle insertion depth in ex vivo porcine eyes are shown in Fig. 6 : For the 300 µm segment (Serial Numbers 1–3), the self-developed Fundus Surgery Image Evaluation Software measured values ranging from 282.6 to 343.9 µm, which were slightly lower overall but fluctuated around the standard line; Kinovea measured values from 389.7 to 673.0 µm, significantly higher than the standard, with a maximum deviation > 120%. For the 400 µm segment (Serial Numbers 4–6), the self-developed Fundus Surgery Image Evaluation Software yielded readings of 409.9–527.4 µm, which were basically consistent with the standard value; Kinovea gave readings of 633.1–1020.8 µm, showing a systematic overestimation with a deviation > 58%. For the 500 µm segment (Serial Numbers 7–9), the self-developed Fundus Surgery Image Evaluation Software measured values of 510.1–836.5 µm; although one measurement was relatively high (836.5 µm), it was still lower than the extreme value of 1287.1 µm measured by Kinovea; Kinovea generally overestimated by 55%–157%. For the 600 µm segment (Serial Numbers 10–12), the self-developed Fundus Surgery Image Evaluation Software obtained values of 505.5–664.8 µm, close to the standard value; Kinovea measured 978.8–1058.9 µm, overestimating by 63%–76%. For the 700 µm segment (Serial Numbers 13–15), the self-developed Fundus Surgery Image Evaluation Software recorded values of 647.4–808.3 µm, slightly lower than or close to the standard; Kinovea continued to overestimate by 33%–69% with values of 931.3–1181.1 µm. For the 800 µm segment (Serial Numbers 16–18), the self-developed Fundus Surgery Image Evaluation Software measured values of 773.1–843.7 µm, fluctuating around the standard line; Kinovea showed one abnormally low value (373.1 µm), while the other values were still overestimated by 37%–38%. For the 900 µm segment (Serial Numbers 19–21), the self-developed Fundus Surgery Image Evaluation Software obtained values of 823.7–1064 µm, slightly lower than or close to the standard; Kinovea had extremely large fluctuations with values of 236.1–1684 µm, including two significant overestimations (1486 µm and 1684 µm). In summary, under the background of ex vivo porcine eyes, the measured needle insertion depths obtained by the self-developed Fundus Surgery Image Evaluation Software were generally consistent with the set standard values, and most deviations were controlled within ± 20%. In contrast, Kinovea software exhibited systematic positive deviations and occasional abnormal underestimations. This indicates that both the measurement stability and accuracy of Kinovea in the background of complex tissues are lower than those of the self-developed Fundus Surgery Image Evaluation Software. The tremor measurement results of ex vivo porcine eyes are shown in Figs. 7 – 8 : For the 300 µm segment (Serial Numbers 1–3): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 6–11 µm and a maximum value of 32–54 µm, both within the low-amplitude range; Kinovea had a mean value of 32–55 µm and a maximum value of 60–335 µm. The mean value of Kinovea was approximately 4–8 times higher than that of the self-developed software, and an extreme peak of 335 µm appeared—indicating that Kinovea misjudged background textures as high-amplitude tremors even in the segment with the lowest standard depth. For the 400 µm segment (Serial Numbers 4–6): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 9–26 µm and a maximum value of 60–103 µm, showing a slight synchronous increase with the increase in depth; Kinovea had a mean value of 47–103 µm and a maximum value of 59–474 µm, with an extreme value of 474 µm, and the deviation further expanded to 3–5 times. For the 500 µm segment (Serial Numbers 7–9): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 9–25 µm and a maximum value of 47–103 µm, with little change in amplitude; Kinovea had a mean value of 0.7–103 µm and a maximum value of 60–159 µm, with severe fluctuations within the segment, indicating poor algorithm stability. For the 600 µm segment (Serial Numbers 10–12): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 9–13 µm and a maximum value of 46–67 µm, maintaining low-amplitude stability; Kinovea had a mean value of 25–68 µm and a maximum value of 29–88 µm, continuing to be systematically 2–3 times higher. For the 700 µm segment (Serial Numbers 13–15): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 7–12 µm and a maximum value of 46–51 µm, with continued stable amplitude; Kinovea had a mean value of 46–67 µm and a maximum value of 45–67 µm. Although the gap with the self-developed software narrowed, it was still 3–6 times higher overall. For the 800 µm segment (Serial Numbers 16–18): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 5–18 µm and a maximum value of 30–155 µm, with a single-point extreme value of 155 µm; Kinovea had a mean value of 1–155 µm and a maximum value of 23–137 µm, with an abnormal low value of 1.7 µm and the largest fluctuations within the segment—indicating that Kinovea had the worst stability when the depth further increased. For the 900 µm segment (Serial Numbers 19–21): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 10–15 µm and a maximum value of 30–61 µm, with amplitude returning to a low level; Kinovea had a mean value of 30–50 µm and a maximum value of 30–61 µm. Although there were no extreme peaks, it was still systematically 2–3 times higher than the self-developed software. In summary, across the seven standard depth segments (300–900 µm), the mean and maximum tremor values measured by the self-developed Fundus Surgery Image Evaluation Software remained < 160 µm, with small differences between segments. This indicates that its algorithm can truly and stably capture the micro-tremor of the instrument against the complex retinal background. In contrast, Kinovea software, due to the sensitivity of its template matching algorithm to vascular textures and reflections, exhibited a systematic overestimation of 2–8 times in all depth segments. It was also accompanied by extreme peaks of 300–400 µm and abnormal low values of 1–2 µm, which demonstrates that Kinovea has a severe deficiency in recognizing tremor signals against complex tissue backgrounds, resulting in low data reliability and inability to meet the requirements of high-precision retinal surgery evaluation.In the tremor comparison of ex vivo porcine eyes, the maximum tremor value of Kinovea for Sample No. 5 (corresponding to the 400 µm segment) was systematically recorded as 35,618.53 µm. This value was two orders of magnitude higher than the other data points in the same group and far exceeded the actual movement range of the instrument. As shown in Fig. 5 , this data point was disconnected from the overall curve, which can be clearly identified as an abnormal peak caused by Kinovea’s template matching algorithm misclassifying large-scale background drift or reflective flicker as "tremor". During subsequent statistical analysis, this outlier was excluded in accordance with the 3σ principle and no longer involved in the calculation of mean values and deviations. 2.3 Processing Results of Clinical Surgical Videos A total of ten clinical surgical videos were selected, and the measurement results of needle insertion depth after analyzing this group of videos are as follows (Fig. 9 ). Depth Range: The measurement results of the self-developed Fundus Surgery Image Evaluation Software ranged from 198 to 934 µm, all falling within the clinically common puncture range (< 1 mm); the measurement results of Kinovea ranged from 174 to 2107 µm, with a span of nearly 12 times, and the maximum value of 2107 µm (Serial Number 3) had exceeded the safe surgical limit. Dispersion and Outliers: The difference between the maximum and minimum measurements of the self-developed Fundus Surgery Image Evaluation Software was 736 µm, with no obvious outliers; the difference between the maximum and minimum measurements of Kinovea was 1932 µm, and an extremely high value of 2107 µm appeared at Serial Number 3, which could be determined as respiratory-ocular movement artifacts misjudged as needle insertion depth. Deviation Trend: Among 8 cases (Serial Numbers 1, 2, 4, 5, 6, 7, 9, 10), the values of Kinovea were generally 50–400 µm higher than those of the self-developed software, indicating a systematic positive deviation; in Serial Number 8, Kinovea showed a slightly lower value than the self-developed Fundus Surgery Image Evaluation Software (174 µm vs. 181 µm), which indicated occasional misalignment between the two algorithms. In summary, among the real surgical videos of manual operation group, the measurement results of the self-developed Fundus Surgery Image Evaluation Software were concentrated in the range of 200–900 µm, which was consistent with clinical operation experience. The self-developed Fundus Surgery Image Evaluation Software could still stably output clinically acceptable depth data against the complex background of living organisms.In contrast, Kinovea could not effectively distinguish between the instrument tip and the moving retina, leading to a significant overestimation of depth. In particular, the abnormal value of 2107 µm at Serial Number 3 had exceeded the safety threshold, which was clinically unacceptable. It can be seen that Kinovea software was interfered by respiration and eye movement, resulting in systematic overestimation and extreme anomalies—indicating that it is not suitable for direct application in the evaluation of in vivo retinal puncture depth. The measurement results of needle insertion tremor in the manual operation group of clinical surgical videos are shown in Figs. 10 – 11 , and the samples with similar performance can be grouped into three categories: low-amplitude stable group (No. 1, 4, 5, 6, 9, 10), medium-amplitude controllable group (No. 2, 7), and extreme abnormal group (No. 3, 8).For the low-amplitude stable group: The self-developed Fundus Surgery Image Evaluation Software measured a tremor mean value of 35–80 µm and a maximum value of 114–252 µm, all within the clinical safety range; Kinovea measured a tremor mean value of 111–161 µm and a maximum value of 295–682 µm, which was overall approximately 1.5–4.6 times higher, showing consistent overestimation but no extreme peaks.For the medium-amplitude controllable group: The self-developed Fundus Surgery Image Evaluation Software measured a tremor mean value of 88–98 µm and a maximum value of 277–384 µm, still lower than the 400 µm warning line; Kinovea measured a tremor mean value of 152–157 µm and a maximum value of 461–507 µm, continuing to be 1.2–1.8 times higher. Although the amplitude was larger, it did not exceed the safety upper limit.For the extreme abnormal group: The self-developed Fundus Surgery Image Evaluation Software measured a tremor mean value of 245 µm (No. 3) and 113 µm (No. 8), and a maximum value of 795 µm (No. 3) and 526 µm (No. 8), still within the acceptable range; in contrast, Kinovea measured a mean value of 577 µm and a maximum value of 2101 µm for No. 3, which had far exceeded the 1 mm safety limit; the maximum value measured for No. 8 by Kinovea was 433 µm, which was less extreme than No. 3 but still significantly higher than the corresponding value of the self-developed software. In summary, among the 10 cases of real surgical videos, 8 cases belong to the low-amplitude stable group and medium-amplitude controllable group, where Kinovea showed systematic overestimation with an amplitude of 1.2–4.6 times. For the 2 cases in the extreme abnormal group, the maximum value of Kinovea for Case No. 3 exceeded 2 mm, which was clearly caused by misjudgment of respiratory-ocular movement drift and clinically unacceptable.The self-developed Fundus Surgery Image Evaluation Software had a mean value < 250 µm and a maximum value < 800 µm in all 10 cases, with concentrated and safe data—demonstrating that it could reliably capture real tremors even against the complex background of living organisms. In contrast, Kinovea could not eliminate background movement, leading to widespread overestimation and occasional extreme anomalies, so it is not suitable for direct application in clinical tremor evaluation. Discussion This study found that in the video processing of in vitro model data, ex vivo porcine eye data, and clinical surgical videos, the values of needle insertion depth and tremor amplitude measured by the Fundus Surgery Image Evaluation Software were closer to the true values than those measured by Kinovea software. These results indicate that the Fundus Surgery Image Evaluation Software can more accurately assess the movement trajectory in delicate operations.Results from the three-stage tests all showed that in terms of needle insertion depth measurement, the data measured by the Fundus Surgery Image Evaluation Software was closer to the true value, while the values measured by Kinovea software were all larger than the set depth values. In terms of tremor amplitude measurement, the tremor values measured by both software were smaller than the set tremor values; however, compared with Kinovea software, the Fundus Surgery Image Evaluation Software was more capable of identifying the tremor amplitude and had better sensitivity. The Fundus Surgery Image Evaluation Software can automatically and more sensitively track the movement of surgical instruments, while Kinovea software cannot automatically identify the key points for tracking nor perform continuous detection. The tracking points of Kinovea software will deviate from the correct positions as the tracking time increases [ 11 ] , thus requiring manual adjustment of the tracking points during long-term tracking [ 12 ] .The Fundus Surgery Image Evaluation Software can automatically identify key points and also support manual adjustment. After the software automatically identifies six points (such as the instrument tip), manual adjustment can be made if there is a deviation. The Fundus Surgery Image Evaluation Software uses the adjusted points as a reference for feedback, thereby achieving continuous self-optimization. This gives the software an advantage in processing video data of human surgical procedures.In ophthalmic surgery, the human eyeball shifts to a certain extent with the patient’s respiratory movement. Against this background, the software needs to track the surgical instruments moving within the vitreous cavity. Background movement will affect the software’s tracking of surgical instruments to a certain extent, leading to inaccurate measurement data. In previous studies, Kinovea software was mostly used in scenarios with a fixed background and the support of a high-definition camera; only under such conditions could Kinovea software maintain a certain level of stability. However, subtle background movements are unavoidable in ophthalmic surgery. Our Fundus Surgery Image Evaluation Software is designed specifically for ophthalmic surgery and has the advantage of handling simultaneous movements of both the surgical background and the surgical instruments.Additionally, the Fundus Surgery Image Evaluation Software achieves more accurate distance conversion. Kinovea software is mostly used in scenarios with camera shooting, while ophthalmic surgeries are performed under a microscope. There is a significant difference in lens distortion between ordinary cameras and microscopes; based on this, we have optimized the Fundus Surgery Image Evaluation Software. With the development of artificial intelligence technology, surgical robots are widely used in the medical field. For subretinal injection, robotic operation has obvious advantages; however, there has been no standard path as a reference for subretinal injection completed by robotic operation in previous studies.The self-developed Fundus Surgery Image Evaluation Software can conduct quantitative analysis on a large number of surgical videos from ophthalmology experts, thereby obtaining the optimal operation path and needle insertion depth for subretinal injection. Meanwhile, it can reduce operational tremor to a certain extent.In the future, the Fundus Surgery Image Evaluation Software can be applied to more surgical operations. By quantifying surgical videos, a relatively standard surgical operation mode can be obtained. Surgical robots can refer to this standard mode to independently complete some simple surgical steps.On the other hand, for surgical beginners, by learning the data analyzed by the Fundus Surgery Image Evaluation Software, they can learn more intuitively and master the key points of some surgical operations more quickly—such as the operation path, needle insertion direction and angle, needle insertion depth, and injection speed of subretinal injection. Declarations The Fundus Surgery Image Evaluation Software is publicly accessible via oculotronics.net to support related research on ophthalmic microsurgery. The limitation of this study lies in that the current version of the Fundus Surgery Image Evaluation Software is only applicable to ophthalmic surgeries. It is adapted to microscopes used in ophthalmic procedures, and there may be certain errors when analyzing videos captured by other types of cameras. Conflict of Interest All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript. Ethical approval This study complied with the Declaration of Helsinki and was approved by the Ethics Committee of the Second Affiliated Hospital of Nanchang University. All participants provided written informed consent. The ex vivo porcine eyes used in this study were obtained from qualified and legitimate abattoirs as by-products of food processing. No animals were sacrificed specifically for this experiment, which complies with the requirements of animal welfare ethics. The experimental protocol has been approved by the Animal Ethics Committee of Nanchang University. Consent to publish Not applicable. Consent to participate Not applicable. Informed consent Informed consent was obtained from all individual participants included in the study. Funding This work was supported by the National Natural Science Foundation of China(No.82360204) and the Jiangxi Provincial Natural Science Foundation General Program (No.20252BAC240512). Availability of data and materials The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. References Del Amo EM, et al. Pharmacokinetic aspects of retinal drug delivery. Prog Retin Eye Res. 2017;57:134–85. Peng Y, Tang L, Zhou Y. Subretinal Injection: A Review on the Novel Route of Therapeutic Delivery for Vitreoretinal Diseases. Ophthalmic Res. 2017;58(4):217–26. Tripepi D et al. The Role of Subretinal Injection in Ophthalmic Surgery: Therapeutic Agent Delivery and Other Indications. Int J Mol Sci, 2023. 24(13). Irigoyen C et al. Subretinal Injection Techniques for Retinal Disease: A Review. J Clin Med, 2022. 11(16). Simunovic MP, et al. Two-step versus 1-step subretinal injection to compare subretinal drug delivery: a randomised study protocol. BMJ Open. 2021;11(12):e049976. Peng J, et al. Subretinal injection of ranibizumab in advanced pediatric vasoproliferative disorders with total retinal detachments. Graefes Arch Clin Exp Ophthalmol. 2020;258(5):1005–12. Jeganathan VS, Shah S. Robotic technology in ophthalmic surgery. Curr Opin Ophthalmol. 2010;21(1):75–80. He B, et al. A review of robotic surgical training: establishing a curriculum and credentialing process in ophthalmology. Eye (Lond). 2021;35(12):3192–201. Roizenblatt M, et al. Robot-assisted tremor control for performance enhancement of retinal microsurgeons. Br J Ophthalmol. 2019;103(8):1195–200. Maberley DAL, et al. A comparison of robotic and manual surgery for internal limiting membrane peeling. Graefes Arch Clin Exp Ophthalmol. 2020;258(4):773–8. Vorapojpisut S, et al. Quantifying sitting posture: A pilot feasibility study of computer vision and wearable sensors (Posture Lab) using a manikin model. Wearable Technol. 2025;6:e27. Garcia-Ruiz P et al. Fiducial Objects: Custom Design and Evaluation. Sens (Basel), 2023. 23(24). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 24 Feb, 2026 Reviews received at journal 23 Feb, 2026 Reviews received at journal 16 Feb, 2026 Reviewers agreed at journal 09 Feb, 2026 Reviewers agreed at journal 03 Feb, 2026 Reviews received at journal 03 Feb, 2026 Reviewers agreed at journal 07 Jan, 2026 Reviewers invited by journal 11 Dec, 2025 Editor assigned by journal 18 Nov, 2025 Submission checks completed at journal 17 Nov, 2025 First submitted to journal 17 Nov, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8029636","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":560426432,"identity":"5e63e453-0e50-4220-abe9-ab976fb019df","order_by":0,"name":"Haonan Xu","email":"","orcid":"","institution":"Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Haonan","middleName":"","lastName":"Xu","suffix":""},{"id":560426434,"identity":"40344141-f8f3-4d62-95c1-14a94a4544dc","order_by":1,"name":"Yunwei Hu","email":"","orcid":"","institution":"Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Yunwei","middleName":"","lastName":"Hu","suffix":""},{"id":560426435,"identity":"1b106fb4-bf7b-4257-812c-d2b44c8456ac","order_by":2,"name":"Yincong Xu","email":"","orcid":"","institution":"The First Hospital of Hebei Medical University，Shijiazhuang","correspondingAuthor":false,"prefix":"","firstName":"Yincong","middleName":"","lastName":"Xu","suffix":""},{"id":560426437,"identity":"0d79ae75-8fa6-4e98-98d4-a98cdb82974a","order_by":3,"name":"Pisong Yan","email":"","orcid":"","institution":"Guangzhou Oculotronics Medical Instrument Co.,LTD","correspondingAuthor":false,"prefix":"","firstName":"Pisong","middleName":"","lastName":"Yan","suffix":""},{"id":560426440,"identity":"69d8ef45-a7e3-4678-9110-0abccecc19f8","order_by":4,"name":"Bowen Zhong","email":"","orcid":"","institution":"Guangzhou Oculotronics Medical Instrument Co.,LTD","correspondingAuthor":false,"prefix":"","firstName":"Bowen","middleName":"","lastName":"Zhong","suffix":""},{"id":560426441,"identity":"3b0a31cb-6d21-4b25-b645-67e0e288297e","order_by":5,"name":"Weifeng Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIie3RsYoCMRCA4ZGFuSbrthEkvkJkYdt7lQSPrRQsU1isrGQLlWt9DMtrDyE2c1dvqfgC2tkp1opZO4t89fwkwwAEwRvCj3J7VhLZ5/Z7v1Nm4k/azCl+NG0BRJHckfMngg9lZ0UihXqInf0savAxRjKNbaanxcgZXSAk1Vx5drHjQWxzXcJ/XuufLnD6W3te2aw3sXXatoqs1oQg+ciTcNUvY3vRiwiysbZRk+Qrba0IU44sg2YJczkcDQrJcMAVOebdpVeVDm6nlL3D7+lsJiKpls+TO+y18SAIguChKy/CSjICDL+fAAAAAElFTkSuQmCC","orcid":"","institution":"Nanchang University","correspondingAuthor":true,"prefix":"","firstName":"Weifeng","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2025-11-04 13:53:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8029636/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8029636/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98777428,"identity":"1fb2b7b8-cccc-4299-9514-68fb969d3d28","added_by":"auto","created_at":"2025-12-22 12:27:04","extension":"tif","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24105226,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1A.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/e4e6bd41e92dc9e39d5eb779.tif"},{"id":98779781,"identity":"216fdf44-8f52-4143-9f80-cd9eca5dbac7","added_by":"auto","created_at":"2025-12-22 12:30:44","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2520360,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/ebf81dfe3659d3014b3108d2.docx"},{"id":98746747,"identity":"9df0b295-3485-4e1d-8ac3-5d2c6a88dc9b","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24105226,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1B.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/a11ed3bc42daa79f16c22367.tif"},{"id":98777645,"identity":"55c77b08-f562-4d93-ad71-53f31fd8dc53","added_by":"auto","created_at":"2025-12-22 12:28:15","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":296608,"visible":true,"origin":"","legend":"","description":"","filename":"Fig10.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/85a35639afc8f79a79692dee.tif"},{"id":98778163,"identity":"5009a8e5-5a86-43b7-85c2-4c70385dd471","added_by":"auto","created_at":"2025-12-22 12:28:57","extension":"tif","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":307210,"visible":true,"origin":"","legend":"","description":"","filename":"Fig11.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/45e6e0cc58975483db20bd09.tif"},{"id":98777340,"identity":"99a3dd22-9850-41e2-b8e0-3f1c3d01944a","added_by":"auto","created_at":"2025-12-22 12:26:34","extension":"tif","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25113866,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2A.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/4ae8f004deeb0584d8396794.tif"},{"id":98777194,"identity":"d52537d7-1b81-44ab-adf8-190c17e655c1","added_by":"auto","created_at":"2025-12-22 12:25:47","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25113866,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2B.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/5a1f42441e8a44ac967df348.tif"},{"id":98778114,"identity":"fd088bc9-4313-4009-bcba-c4e0d26aca3c","added_by":"auto","created_at":"2025-12-22 12:28:55","extension":"tif","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":281662,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3A.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/44928c4a580c12a314666373.tif"},{"id":98779045,"identity":"663ea497-bf6f-450f-ba4a-43a2c712bde3","added_by":"auto","created_at":"2025-12-22 12:29:53","extension":"tif","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":265980,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3B.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/bce106b60fb84af874d0262a.tif"},{"id":98777332,"identity":"4b1e2599-89b2-444d-a589-7acba1048efc","added_by":"auto","created_at":"2025-12-22 12:26:33","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":385600,"visible":true,"origin":"","legend":"","description":"","filename":"Fig4.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/39d122eef229324c17bbaaa9.tif"},{"id":98777098,"identity":"00ef6a5c-d9aa-4925-9ae6-f2114053a637","added_by":"auto","created_at":"2025-12-22 12:25:21","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":364254,"visible":true,"origin":"","legend":"","description":"","filename":"Fig5.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/43390548b1e32c6fe62195e8.tif"},{"id":98746763,"identity":"191d0eee-8c8b-4d8c-b4f2-e113cd30e20c","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"tif","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":373020,"visible":true,"origin":"","legend":"","description":"","filename":"Fig6.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/604bc912036572218871c650.tif"},{"id":98777075,"identity":"e1444e61-a57b-4fa1-bf1d-2d1d64b7d943","added_by":"auto","created_at":"2025-12-22 12:25:19","extension":"tif","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":315302,"visible":true,"origin":"","legend":"","description":"","filename":"Fig7.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/f49c83fecd47c08f13692c35.tif"},{"id":98779066,"identity":"b68929d5-baad-4cb2-9d55-1759a595a705","added_by":"auto","created_at":"2025-12-22 12:29:56","extension":"tif","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313346,"visible":true,"origin":"","legend":"","description":"","filename":"Fig8.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/cab066684794db91bce7bf8d.tif"},{"id":98777111,"identity":"2dcdabde-0e4e-43b2-a79c-dcd4f16c68c6","added_by":"auto","created_at":"2025-12-22 12:25:24","extension":"tif","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313828,"visible":true,"origin":"","legend":"","description":"","filename":"Fig9.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/62bdf3b070c6a0895f7e399f.tif"},{"id":98778546,"identity":"6a0e7dbc-6733-4d5a-b59c-2f9d4904f855","added_by":"auto","created_at":"2025-12-22 12:29:26","extension":"json","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9028,"visible":true,"origin":"","legend":"","description":"","filename":"d5d3be3e509d4abb96578e32337b0720.json","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/bfae5100bac0d40ee0cfab84.json"},{"id":98778368,"identity":"9aafb46e-0653-4483-8539-aeea6c427c68","added_by":"auto","created_at":"2025-12-22 12:29:12","extension":"xml","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95054,"visible":true,"origin":"","legend":"","description":"","filename":"d5d3be3e509d4abb96578e32337b07201enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/c6c2b40021bfc32540f94322.xml"},{"id":98779854,"identity":"f018cede-840e-4119-b255-ee8e3d926d86","added_by":"auto","created_at":"2025-12-22 12:30:50","extension":"tif","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24105226,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1A.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/2b31229ee623fdb843618880.tif"},{"id":98746783,"identity":"9d1b313b-ff48-4312-bed4-b265a2fd2f0e","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"tif","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24105226,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1B.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/8260d4df919f941881aea91e.tif"},{"id":98777127,"identity":"2e49045b-099b-4dab-9c8a-336e1494353c","added_by":"auto","created_at":"2025-12-22 12:25:27","extension":"tif","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":296608,"visible":true,"origin":"","legend":"","description":"","filename":"Fig10.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/c03598b5fb27d83b14553be4.tif"},{"id":98746781,"identity":"9faf4596-38b8-4e51-b8db-3d56c67547d8","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"tif","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":307210,"visible":true,"origin":"","legend":"","description":"","filename":"Fig11.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/b45c1132abaf37b38ce8d353.tif"},{"id":98777010,"identity":"95ceeb20-51d6-430b-9f75-0e66b7e87401","added_by":"auto","created_at":"2025-12-22 12:25:01","extension":"tif","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25113866,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2A.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/ae6ad0588bbdf713ef38f4cd.tif"},{"id":98777468,"identity":"0911477f-9db2-484c-b205-e8cc9af58d17","added_by":"auto","created_at":"2025-12-22 12:27:27","extension":"tif","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25113866,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2B.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/b86b26eb6efe24ef21b53253.tif"},{"id":98778603,"identity":"080d5214-9d31-434a-80c8-1479de138826","added_by":"auto","created_at":"2025-12-22 12:29:28","extension":"tif","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":281662,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3A.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/222faabbe39c5ba749a6a668.tif"},{"id":98746807,"identity":"e0e45d10-f730-4775-9e13-c91eae198df8","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"tif","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":265980,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3B.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/67375a405dbfd1228f24fa0c.tif"},{"id":98779808,"identity":"3336a0e7-aa33-4565-a711-5b1e4d6834b4","added_by":"auto","created_at":"2025-12-22 12:30:47","extension":"tif","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":385600,"visible":true,"origin":"","legend":"","description":"","filename":"Fig4.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/a05f851bb5f3044a2e8814ee.tif"},{"id":98778073,"identity":"5b6b3252-606b-455a-af50-b2fb209a5b18","added_by":"auto","created_at":"2025-12-22 12:28:52","extension":"tif","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":364254,"visible":true,"origin":"","legend":"","description":"","filename":"Fig5.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/2fe7ff6cd2ea195b18e72ccc.tif"},{"id":98746776,"identity":"37c81165-112f-4fb4-ba53-bc821ed853c8","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"tif","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":373020,"visible":true,"origin":"","legend":"","description":"","filename":"Fig6.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/3d5a37f67a61c0c2b7a7638e.tif"},{"id":98778925,"identity":"0ae47788-c9d6-42c8-a985-36d7e4640b9f","added_by":"auto","created_at":"2025-12-22 12:29:49","extension":"tif","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":315302,"visible":true,"origin":"","legend":"","description":"","filename":"Fig7.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/c50ce5ea8539817341b3acd3.tif"},{"id":98777453,"identity":"c757b312-b174-4ccd-8927-4cc6781131ca","added_by":"auto","created_at":"2025-12-22 12:27:26","extension":"tif","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313346,"visible":true,"origin":"","legend":"","description":"","filename":"Fig8.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/ce49a0c9fc08d6574c914fd0.tif"},{"id":98746786,"identity":"7d834e00-fbc7-474d-8ca9-87c94e5c17bd","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"tif","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313828,"visible":true,"origin":"","legend":"","description":"","filename":"Fig9.tif","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/514ae34beea1161cfed699f9.tif"},{"id":98797639,"identity":"c3f7509c-0dd6-42df-a7fd-115d08a18c54","added_by":"auto","created_at":"2025-12-22 13:38:48","extension":"jpeg","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":215796,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/e8f885460f7adb69575b1090.jpeg"},{"id":98746761,"identity":"85371652-a430-4ad6-8a55-34a880e2e326","added_by":"auto","created_at":"2025-12-22 08:47:57","extension":"jpeg","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":296608,"visible":true,"origin":"","legend":"","description":"","filename":"Fig10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/0dfc5b00048e7a1e1b132e82.jpeg"},{"id":98777392,"identity":"b0c85749-6d2c-466a-90f3-402adb64993c","added_by":"auto","created_at":"2025-12-22 12:26:46","extension":"jpeg","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":307210,"visible":true,"origin":"","legend":"","description":"","filename":"Fig11.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/0243867f8312dc015edd7741.jpeg"},{"id":98778418,"identity":"af2de6bd-c238-44bd-9c10-37a8638392c7","added_by":"auto","created_at":"2025-12-22 12:29:14","extension":"jpeg","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":201953,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/d5bced4d1ff34ded9bfd2033.jpeg"},{"id":98778323,"identity":"4beebf18-c35f-479a-b418-6f77dfce69f1","added_by":"auto","created_at":"2025-12-22 12:29:10","extension":"jpeg","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":150440,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/56a80bc267ff4df5736fdf5d.jpeg"},{"id":98778342,"identity":"a11c05dd-0010-455c-b448-001102be4c5a","added_by":"auto","created_at":"2025-12-22 12:29:11","extension":"jpeg","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":385600,"visible":true,"origin":"","legend":"","description":"","filename":"Fig4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/486acae64e4e2bef6946d69c.jpeg"},{"id":98776989,"identity":"1ffcad8d-58fd-42ed-aba4-60ca7313bb2b","added_by":"auto","created_at":"2025-12-22 12:24:54","extension":"jpeg","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":364254,"visible":true,"origin":"","legend":"","description":"","filename":"Fig5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/2deb09c4bc801478b4740056.jpeg"},{"id":98746768,"identity":"bd86e2d0-7fb3-407e-8f90-7eb632796d4d","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"jpeg","order_by":38,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":373020,"visible":true,"origin":"","legend":"","description":"","filename":"Fig6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/e888e9aec54e6db8441bb286.jpeg"},{"id":98779693,"identity":"473ef252-f0a2-400d-9831-50b6cec2be2a","added_by":"auto","created_at":"2025-12-22 12:30:36","extension":"jpeg","order_by":39,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":315302,"visible":true,"origin":"","legend":"","description":"","filename":"Fig7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/429e1dc15c6253c3f017d934.jpeg"},{"id":98746769,"identity":"adf67de2-6a55-422c-b5ac-d4779290d179","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"jpeg","order_by":40,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313346,"visible":true,"origin":"","legend":"","description":"","filename":"Fig8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/591f3f6aba378cf30413d18c.jpeg"},{"id":98778592,"identity":"2e942a0c-b019-4996-9eb8-353add704d0f","added_by":"auto","created_at":"2025-12-22 12:29:27","extension":"jpeg","order_by":41,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313828,"visible":true,"origin":"","legend":"","description":"","filename":"Fig9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/8ca89b45e7e04ebd7227be54.jpeg"},{"id":98779085,"identity":"7e5140f3-461f-4327-8e27-f619676e66a7","added_by":"auto","created_at":"2025-12-22 12:29:56","extension":"png","order_by":42,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":298551,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig1A.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/d7708f4d4cff90677a88f657.png"},{"id":98777169,"identity":"6e546e78-67c1-4f8f-8d58-54697371e149","added_by":"auto","created_at":"2025-12-22 12:25:32","extension":"png","order_by":43,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":101157,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig1B.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/701f171b102c9618432cb952.png"},{"id":98777377,"identity":"b05e5069-5c6e-42ce-bf40-79c54bfb4864","added_by":"auto","created_at":"2025-12-22 12:26:37","extension":"png","order_by":44,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24445,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig10.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/b80f146ece089653ec2dd0bd.png"},{"id":98777181,"identity":"43e7557c-a154-40fd-a48b-c8c3880cc9d7","added_by":"auto","created_at":"2025-12-22 12:25:44","extension":"png","order_by":45,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25841,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig11.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/0a82a227d40dc150c9855f73.png"},{"id":98746806,"identity":"237e4937-d850-4ec9-98ef-98bd60f7bb6c","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":46,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":124441,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig2A.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/00ed54e39c6f5ef08c3bcaed.png"},{"id":98746766,"identity":"73e5467b-8167-4b4e-b807-e7f54b29574b","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"png","order_by":47,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":124212,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig2B.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/4327b2f8b99f10359d08b960.png"},{"id":98778167,"identity":"6ecbd40a-adae-4549-bb53-3cf2950c170b","added_by":"auto","created_at":"2025-12-22 12:28:57","extension":"png","order_by":48,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27940,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig3A.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/bc886661bd95f46a618df296.png"},{"id":98746794,"identity":"ab3d85e9-53c7-426c-b28a-2b32c2e74c2b","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":49,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":26878,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig3B.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/5349337d42561b8dfbab2258.png"},{"id":98777100,"identity":"18253382-dd98-4b8a-a8c4-1e3a982bdf69","added_by":"auto","created_at":"2025-12-22 12:25:22","extension":"png","order_by":50,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35106,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig4.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/6e9d1774482b654f1c52c39b.png"},{"id":98778105,"identity":"184006ef-49c0-43dc-bec0-565d74b08e00","added_by":"auto","created_at":"2025-12-22 12:28:55","extension":"png","order_by":51,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36175,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig5.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/1fa08c7578b2b080c21f2472.png"},{"id":98779036,"identity":"c8a827dc-8266-43af-accd-4f1839ab467e","added_by":"auto","created_at":"2025-12-22 12:29:53","extension":"png","order_by":52,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36354,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig6.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/f47f1f992cb5015b8b511ad2.png"},{"id":98777188,"identity":"eafbdb63-1361-4b67-aa9f-e56e778d292c","added_by":"auto","created_at":"2025-12-22 12:25:47","extension":"png","order_by":53,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28152,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig7.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/2db7ad9df2b92750d9c8bd68.png"},{"id":98746803,"identity":"6a497ba8-02e5-470e-9427-f88b7e114810","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":54,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27377,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig8.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/71a31158f561ac873703ab9b.png"},{"id":98746795,"identity":"a82778fc-2588-4fec-8bb2-47dbfc3605cf","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":55,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27913,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig9.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/99583dc36abf1c135024bd67.png"},{"id":98777148,"identity":"6499b045-7af2-4fd9-836b-2b4c54df207c","added_by":"auto","created_at":"2025-12-22 12:25:31","extension":"png","order_by":56,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53061,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/995fbd13224ae30ee991bc92.png"},{"id":98778120,"identity":"37b1079a-4a5b-449a-80ac-da7f2dac9078","added_by":"auto","created_at":"2025-12-22 12:28:55","extension":"png","order_by":57,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24445,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/5af3d8eff8e9a47f3ff5b751.png"},{"id":98777295,"identity":"a553c14d-90d3-489f-8cea-78289c67b7e0","added_by":"auto","created_at":"2025-12-22 12:26:30","extension":"png","order_by":58,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25841,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/9fe81106832cb320fb3622a0.png"},{"id":98779113,"identity":"4f35a621-3d60-4fec-ae5c-1c28cb100011","added_by":"auto","created_at":"2025-12-22 12:29:58","extension":"png","order_by":59,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43782,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/d8abe8e518747e76167b585b.png"},{"id":98746800,"identity":"f4166e6c-3a0d-4a2a-8bba-278304b36574","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":60,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":45119,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/cd683d8a92d53d0cd98f9a84.png"},{"id":98779089,"identity":"d4b9c823-b461-4d52-9767-284f091d38ad","added_by":"auto","created_at":"2025-12-22 12:29:57","extension":"png","order_by":61,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35106,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/8534a53b0c7c6e030a3c4f13.png"},{"id":98746792,"identity":"02e467f2-cd9a-44c0-a282-036dbc402723","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":62,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36175,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/2d71b221e8ebc43f2d5b784c.png"},{"id":98779052,"identity":"b0eb0ec0-8569-4c9e-8b79-84d67e89a90b","added_by":"auto","created_at":"2025-12-22 12:29:55","extension":"png","order_by":63,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36354,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/700e25b496b3a5b05218991b.png"},{"id":98746805,"identity":"d318fa05-dc73-4d9e-b401-198d6e84cfb5","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":64,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28152,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/3fc714e3e86c4ded16b754e5.png"},{"id":98746798,"identity":"0568a133-c91b-4f02-8eb9-bc57d4768f33","added_by":"auto","created_at":"2025-12-22 08:47:59","extension":"png","order_by":65,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27377,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/adee27098c51eb0ea96ddcc5.png"},{"id":98746767,"identity":"433da504-01b1-4782-b991-4daccfdac64b","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"png","order_by":66,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27913,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/d6189778cd4f0161a40f2304.png"},{"id":98777921,"identity":"90da1f5c-11d5-4548-bcd3-d109e8252db8","added_by":"auto","created_at":"2025-12-22 12:28:40","extension":"xml","order_by":67,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":88896,"visible":true,"origin":"","legend":"","description":"","filename":"d5d3be3e509d4abb96578e32337b07201structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/2ab53319d53895625604d0f8.xml"},{"id":98746771,"identity":"9506cbe8-fdc9-4b84-97e9-a7a2296ee675","added_by":"auto","created_at":"2025-12-22 08:47:58","extension":"html","order_by":68,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":104416,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/997778c0228b07920df2cd5e.html"},{"id":98746730,"identity":"bf4a0ecc-ed4b-4fc8-85e1-e3b0b89ea091","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":328673,"visible":true,"origin":"","legend":"\u003cp\u003e(A)shows the positioning points (red, green, and yellow) used by the Fundus Surgery image evaluation software for measuring needle insertion depth against a blank background.(B) shows the positioning points (red, green, and yellow) used by the Fundus surgery image evaluation software for measuring tremor values against a blank background.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/3341d83d65073e403fd37c46.png"},{"id":98746729,"identity":"7872beb3-61e4-406b-8f3e-6c7afd0ad900","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":296074,"visible":true,"origin":"","legend":"\u003cp\u003e(A) shows the positioning points (red, green, and yellow) used by the Fundus Surgery Image Evaluation Software for measuring needle insertion depth and tremor during the operation against an ex vivo porcine eye background. (B) shows the measurement of needle insertion depth and tremor during the operation (which were originally measured by the Fundus Surgery Image Evaluation Software against an ex vivo porcine eye background) using the Kinovea software against a blank background.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/9b559bed6de26499b34688f7.png"},{"id":98746728,"identity":"e86a16f0-4e03-4b35-a4e6-0256943ee2ed","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":287974,"visible":true,"origin":"","legend":"\u003cp\u003e(A) shows that both the video points and axis analyzed by the Fundus Surgery Image Evaluation Software can be accurately identified against a human retinal background. (B)shows that the tracking frame of the Kinovea software drifts under poor video quality conditions against a human retinal background, resulting in abnormal needle insertion tremor results.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/603f470859cbe296593466f4.png"},{"id":98779065,"identity":"f5ac574a-fcf1-43fd-a8f2-fde9fa1eb7ae","added_by":"auto","created_at":"2025-12-22 12:29:56","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":217316,"visible":true,"origin":"","legend":"\u003cp\u003eMeasurement Results of Needle Insertion Depth in the In Vitro Model: In the figure, the blue dots represent the measurement results of the self-developed Fundus Surgery Image Evaluation Software, the orange triangles represent the measurement results of Kinovea software, and the blue line represents the set standard needle insertion depth. The abscissa denotes the serial numbers (1–25) of the 25 repeated measurements, and the ordinate represents the measured needle insertion depth (μm).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/f8dc40a6c554621ddc83c7ec.png"},{"id":98777505,"identity":"313dd9ef-0d28-452d-b703-0fa48d0ada71","added_by":"auto","created_at":"2025-12-22 12:27:51","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":291514,"visible":true,"origin":"","legend":"\u003cp\u003eMeasurement Results of Tremor Amplitude in the In Vitro Model: In the figure, the orange color represents the mean tremor values measured by the self-developed Fundus Surgery Image Evaluation Software, and the gray color represents its maximum tremor values; the yellow color represents the mean tremor values of Kinovea, and the light blue triangles represent its maximum tremor values; the blue horizontal lines indicate the set standard tremor values (50, 100, 150, 200, 250 μm). The abscissa denotes the serial numbers (1–25) of the 25 repeated measurements, and the ordinate represents the measured tremor amplitude (μm).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/c359db006da85e4c9b5b4ce2.png"},{"id":98746733,"identity":"e776fc54-a694-4a3a-bbaf-36fd03aa8412","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":222011,"visible":true,"origin":"","legend":"\u003cp\u003eMeasurement Results of Needle Insertion Depth in Ex Vivo Porcine Eyes: In the figure, the blue line represents the measured values of needle insertion depth from the self-developed Fundus Surgery Image Evaluation Software, the gray color represents the measured values from Kinovea software, and the [reference] line denotes the set standard needle insertion depth (300–900 μm). The abscissa indicates the serial numbers (1–21) of the 21 repeated measurements, and the ordinate represents the measured needle insertion depth (μm).\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/c0cb2d19442157bcdb87694d.png"},{"id":98777207,"identity":"bf932dc6-f006-413f-9b45-3ab9f8a4ebfc","added_by":"auto","created_at":"2025-12-22 12:25:53","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":170004,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Mean Tremor Values in Ex Vivo Porcine Eyes: In the figure, the blue bar charts represent the mean tremor values measured by the self-developed Fundus Surgery Image Evaluation Software, and the orange bar charts represent the mean tremor values of Kinovea; the abscissa denotes the serial numbers (1–21) of the 21 repeated measurements, corresponding to the set standard needle insertion depths (300 μm, 400 μm, 500 μm, 600 μm, 700 μm, 800 μm, 900 μm), with 3 data points for each depth level; the ordinate represents the measured tremor amplitude (μm).\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/be38fce9a5a1b62b14829da9.png"},{"id":98746741,"identity":"3f47eba3-39ea-482b-bb2a-fc0e4a4366b3","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":182058,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Maximum Tremor Values in Ex Vivo Porcine Eyes: In the figure, the blue bar charts represent the maximum tremor values measured by the self-developed Fundus Surgery Image Evaluation Software, and the [orange] bar charts represent the maximum tremor values of Kinovea; the abscissa denotes the serial numbers (1–21) of the 21 repeated measurements, corresponding to the set standard needle insertion depths (300 μm, 400 μm, 500 μm, 600 μm, 700 μm, 800 μm, 900 μm), with 3 data points for each depth level; the ordinate represents the measured tremor amplitude (μm).\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/e4585e613098c1be5216e06a.png"},{"id":98779010,"identity":"eaf79df6-ae13-4806-bf96-2f7b925783a7","added_by":"auto","created_at":"2025-12-22 12:29:52","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":181304,"visible":true,"origin":"","legend":"\u003cp\u003eMeasurement Results of Needle Insertion Depth in the Manual Operation Group of Clinical Surgical Videos: In the figure, the blue bar charts represent the measured values of the self-developed Fundus Surgery Image Evaluation Software, and the orange bar charts represent the measured values of Kinovea; the abscissa denotes the serial numbers (1–10) of the 10 clinical surgical videos, and the ordinate represents the measured needle insertion depth (μm). The background is the moving retina of a living organism, with interferences from respiration and slight eye movement.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/b14d565699066f4d5cee0fb3.png"},{"id":98746739,"identity":"4cebd848-8768-49cf-9802-e8b8ddb76b92","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":163532,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Mean Tremor Values in the Manual Operation Group of Clinical Surgical Videos: The blue bar charts represent the mean tremor values of the self-developed Fundus Surgery Image Evaluation Software, and the orange bar charts represent the mean tremor values of Kinovea; the blue bar charts in Figure 8 represent the maximum tremor values of the self-developed Fundus Surgery Image Evaluation Software, and the orange bar charts represent the maximum tremor values of Kinovea. The abscissa denotes the serial numbers (1–10) of the 10 clinical surgical videos, and the ordinates represent the mean tremor amplitude (μm) and maximum tremor amplitude (μm) respectively. The background is the moving retina of a living organism, with interferences from respiration and slight eye movement.\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/85285545f5d30f9534ce837d.png"},{"id":98746734,"identity":"e8f6ab67-3281-472a-8200-4b7a46033d5a","added_by":"auto","created_at":"2025-12-22 08:47:56","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":168360,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Maximum Tremor Values in the Manual Operation Group of Clinical Surgical Videos: In the figure, the blue bar charts represent the maximum tremor values measured by the self-developed Fundus Surgery Image Evaluation Software, and the orange bar charts represent the maximum tremor values of Kinovea; the abscissa denotes the serial numbers (1–10) of the 10 clinical surgical videos, and the ordinates represent the mean tremor amplitude (μm) and maximum tremor amplitude (μm) respectively. The background is the moving retina of a living organism, with interferences from respiration and slight eye movement.\u003c/p\u003e","description":"","filename":"11.png","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/eec8566d55ce208bea309e6a.png"},{"id":98799659,"identity":"381f67ba-6de5-43ec-9aba-fde30b2b370d","added_by":"auto","created_at":"2025-12-22 14:13:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3848001,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8029636/v1/580aba71-3033-4fe5-9a13-81cf715df066.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Analysis of Performance Differences Between Self-Developed Fundus Surgery Image Evaluation Software and Kinovea in Ophthalmic Microsurgery","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSubretinal injection refers to the direct delivery of drugs into the potential space between the neurosensory retina and the retinal pigment epithelium (RPE) layer, namely the subretinal space. It is an administration method that delivers therapeutic drugs to the subretinal space\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. Currently, the most common administration route for fundus diseases remains intravitreal injection, where drugs are directly injected into the vitreous humor, enabling rapid achievement of high concentrations. However, drugs may be cleared relatively quickly from the vitreous humor, necessitating frequent administrations. Compared with intravitreal injection, subretinal injection\u0026mdash;an administration method that has developed rapidly in recent years\u0026mdash;has become increasingly prevalent in ophthalmic clinical practice and research. It offers advantages including direct drug action, high local drug concentration, and prolonged duration of interaction with the retina\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e, while minimizing the risks of immune rejection and inflammation\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSubretinal injection requires the surgeon to accurately penetrate the neurosensory retina (with a thickness of only 150\u0026ndash;200 \u0026micro;m) using an injection needle without damaging the retinal pigment epithelium (RPE) layer\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e. Additionally, it is necessary to ensure a certain level of stability in needle insertion depth, injection volume, and injection pressure to prevent further retinal damage\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. These requirements make the technique highly demanding on the surgeon\u0026rsquo;s operational skills, highlighting the critical importance of a dedicated device capable of maintaining excellent stability.\u003c/p\u003e \u003cp\u003eSurgical robots can to a certain extent address the issue that it is difficult for surgeons to maintain a high standard of stability during subretinal injection. Combined with microsurgical technology, surgical robots can assist ophthalmologists in performing more precise operations during surgery, maintain operational stability, and simultaneously reduce secondary damage to normal tissues during the surgical process\u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. Currently, ophthalmic surgical robots that have been applied in clinical practice have been proven to exhibit less tremor compared with manual operations during retinal surgery; they can overcome the physiological limitation of surgeon fatigue and maintain the stability of surgical instruments\u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e. However, we have found that there is a lack of quantitative analysis on the operational path of surgical robots, and there is also an absence of a relatively standard operational model to serve as a comparison benchmark.\u003c/p\u003e \u003cp\u003eAgainst this background, for beginners in ophthalmic surgery, the traditional training model lacks a quantitative evaluation system, making it difficult to help them accurately master the key points of surgery. In the field of surgical robot research and development, a standardized operational model is also required to improve the accuracy of automated systems. With the in-depth application of artificial intelligence (AI) technology in the medical field, quantitative analysis of surgical videos has become a key approach to solving the aforementioned problems. On the one hand, by leveraging computer vision and AI technologies to extract spatiotemporal features and analyze motion sequences from expert surgical videos, complex surgical operations can be converted into quantifiable parameter indicators. This provides beginners with intuitive learning feedback and improvement directions, accelerating the process of skill acquisition. On the other hand, these quantitative analysis results can also provide surgical robots with references for standard operating methods and precise path planning, facilitating the construction of more intelligent and safer automated surgical assistance systems and promoting the advancement of subretinal injection technology toward standardization and intelligentization.\u003c/p\u003e \u003cp\u003eCurrently, we have observed that Kinovea software is used in rehabilitation medicine, sports medicine, biomechanical research, and clinical evaluation to analyze motion trajectories and force conditions. Kinovea is an open-source video analysis software (specializing in 2D motion tracking and biomechanical analysis), primarily applicable to basic video analysis and sports biomechanical analysis. It has been widely adopted due to its free and open-source nature, low learning threshold, and relatively professional biomechanical analysis capabilities. However, its limitations\u0026mdash;including limited precision, low automation, lack of 3D analysis, and insufficient medical adaptation\u0026mdash;indicate that it is unsuitable for high-precision ophthalmic surgeries. To date, no team has developed a software specifically tailored for ophthalmic surgical procedures. Therefore, a software capable of analyzing high-precision ophthalmic surgeries and capturing surgical movements for quantitative analysis is particularly crucial.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e1.1 Software System\u003c/h2\u003e\n \u003cp\u003eThe self-developed Fundus Surgery Image Evaluation Software is a high-precision analysis tool specifically designed for ophthalmic surgeries. Its primary function is to quantify key parameters during ophthalmic surgical procedures, such as needle insertion depth measurement and tremor magnitude, thereby providing data support for analyzing the accuracy of surgical operations.\u003c/p\u003e\n \u003cp\u003eThe Fundus Surgery Image Evaluation Software consists of four core modules: 1. Video Selection: Including rough clipping, surgical video selection, playback, pause, and video segment clipping; 2. Pre-analysis: Encompassing data annotation and fine adjustment of marked points; 3. Reference Frame: Involving image selection and reference frame verification; 4. Data Analysis: Covering predicted point visualization, indicator data visualization, and report output.\u003c/p\u003e\n \u003cp\u003eThe basic operational workflow is as follows:\u003c/p\u003e\n \u003cp\u003e\u003cspan\u003e1. Rough Clipping: Extract key steps from the entire surgical video, including sequential operations such as pre-insertion, injection, fundus elevation, and post-withdrawal, while excluding invalid frames. Drag the surgical video from the folder into the editing software project file and add it to the timeline. Confirm the video segment by dragging the scale bar, click the \u0026quot;Clip\u0026quot; button to perform clipping, select and delete unwanted clipped segments, and finally click the \u0026quot;Export\u0026quot; button in the toolbar to select the output video format and save path.\u003cbr\u003e\u003c/span\u003e \u003cspan\u003e2. Video Selection: Click \u0026quot;Select Video Source\u0026quot; to choose and open the surgical video. Then click to select the corresponding video type, and mark whether the current video is recorded from manual operation or robotic operation.\u003cbr\u003e\u003c/span\u003e \u003cspan\u003e3. Video Frame Selection: Determine the standard positions of the left and right selection frames based on surgical movements. For example, in the \u0026quot;subretinal injection evaluation\u0026quot; mode, the left and right frame ranges can be defined using the keyframe method, by dragging the progress bar to record the left and right frame indices, or by manually entering the indices for frame selection. Click the \u0026quot;Pre-analysis\u0026quot; button after completing frame selection.\u003cbr\u003e\u003c/span\u003e \u003cspan\u003e4. Pre-analysis: Annotate the microneedle point on the instrument detection frame. Click \u0026quot;Open Schematic\u0026quot; to view the position of the marked point, switch between images using the \u0026quot;Previous\u0026quot; and \u0026quot;Next\u0026quot; buttons, and click \u0026quot;Start Calibration\u0026quot; to calibrate the position. The image can be zoomed in or out using the shortcut \u0026quot;Ctrl\u0026thinsp;+\u0026thinsp;Mouse Wheel\u0026quot;. Click \u0026quot;Start Pre-analysis\u0026quot; after calibration is completed; if there are redundant selection frames or missing marked points, corresponding adjustments can be made.\u003cbr\u003e\u003c/span\u003e \u003cspan\u003e5. Reference Frame: Drag an image that meets the criteria of complete and clear needle tip, and close proximity to the retina, to set it as the reference frame. Click \u0026quot;Verify Reference Frame\u0026quot; to perform verification; the image can be zoomed using the mouse wheel. After confirming that the image meets the criteria, check the selection box and click \u0026quot;Verify\u0026quot;.\u003cbr\u003e\u003c/span\u003e \u003cspan\u003e6. Data Analysis: Click \u0026quot;Start Analysis\u0026quot; to initiate the analysis process, and click \u0026quot;Stop Analysis\u0026quot; to terminate it if needed. During the analysis, visualization of predicted points and evaluation indicators can be performed. Click the \u0026quot;Play\u0026quot; button to control the playback and pause of the data video, and drag the progress bar to control the display progress. Finally, click \u0026quot;Export Report\u0026quot;, fill in the required information, and click \u0026quot;Print\u0026quot; or \u0026quot;Save\u0026quot; to complete the report output.\u003cbr\u003e\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eKinovea is a feature-rich and powerful open-source tool for video recording, measurement, and annotation. It supports image visualization, recording, measurement, annotation, and (high-speed) motion analysis, enabling users to capture, compare, annotate, and measure movements in videos. It has a wide range of application scenarios, including sports, laboratory settings, and industrial automation.\u003c/p\u003e\n \u003cp\u003eThe basic operational workflow is as follows:\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e1. Image or Video Acquisition\u003c/h3\u003e\n\u003cp\u003eLaunch the Kinovea software, locate the camera name in the software interface and double-click it to obtain the real-time image from the camera. To import a video from a file, you can use File Explorer to navigate to the folder where the video is stored, then double-click the thumbnail of the video file in the thumbnail panel on the right to open the video in the player screen; alternatively, you can use the \"File\u0026thinsp;\u0026gt;\u0026thinsp;Open\" menu option, or directly drag the video file from Windows Explorer to the Kinovea software interface to open it.\u003c/p\u003e\n\u003ch3\u003e2. Video Operations\u003c/h3\u003e\n\u003cp\u003e(1) Playback Control: After opening the video, use the control buttons of the player to perform operations such as starting playback, pausing, fast-forwarding, and rewinding; you can also drag the navigation cursor to jump to any position in the video.\u003c/p\u003e \u003cp\u003e(2) Workspace Setup: Navigate to the start position of the segment you are interested in within the video and click the \"Workspace Start\" button; then play the video until the end of the segment requiring analysis, and click the \"Workspace End\" button. After setting up the workspace, if the loop playback mode is enabled, the video will play in a loop within the workspace, and the navigation cursor will be more precise within the workspace at this time.\u003c/p\u003e \u003cp\u003e(3) Image Size Adjustment: If the video image is too small, you can drag the small squares at the corners of the image to adjust its size; alternatively, you can directly perform zoom operations on the image or use the magnifier tool to enlarge image details.\u003c/p\u003e \u003cp\u003e(4) Playback Speed Adjustment: If a more detailed study of movements is needed, you can reduce the playback speed using the speed slider. To restore normal speed (100%), simply double-click the speed percentage value. You can also use the up and down arrow keys on the keyboard to adjust the speed in small increments.\u003c/p\u003e\n\u003ch3\u003e3. Image and Video Annotation\u003c/h3\u003e\n\u003cp\u003e(1) Distance Measurement: Click the line button in the annotation tools and draw a line on the image. After drawing, left-click the line and select \"Show Measurements\u0026thinsp;\u0026gt;\u0026thinsp;Length\", at which point the pixel length of the line will be displayed. To obtain accurate measurements in actual millimeters, it is necessary to place an object of known size (e.g., a ruler) in the image. Draw a line corresponding to the known actual length, left-click the line, select \"Calibrate\", enter the actual length of the line, and click \"Apply\". All subsequent lines drawn will display their actual lengths based on this calibration.\u003c/p\u003e \u003cp\u003e(2) Adding Annotations: Kinovea provides a variety of annotation tools, such as those for adding arrows, text descriptions, and drawing shapes, which are used to highlight key positions or features in the video. Simply click the corresponding annotation tool button and draw or add content directly on the video screen.\u003c/p\u003e \u003cp\u003e(3) Adding Key Images: When the video plays to a key position, click the \"Add Key Image\" button to mark the current frame as a key image. Thumbnails of key images will be displayed in the software interface, facilitating subsequent review and comparison.\u003c/p\u003e\n\u003ch3\u003e4. Data Export and Saving\u003c/h3\u003e\n\u003cp\u003eAfter completing video analysis and annotation, the analysis data can be exported to spreadsheet formats (e.g., OpenOffice Calc, Excel, or plain text files) for further processing and analysis. Through the corresponding export menu options, select the type of data to be exported (e.g., time, length, angle values) to perform the export. If it is necessary to save the video or its key images, select the \"Save\" option via the \"File\" menu, and choose the save format and path according to the prompts. For the saved video, if there are drawn and annotated contents during the analysis process, you can select whether to embed these contents into the video based on the software settings.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e1.2 Validation Experiment Design\u003c/h2\u003e \u003cp\u003eThis study systematically evaluated the performance of the self-developed Fundus Surgery Image Evaluation Software and the conventional Kinovea software in measuring ophthalmic surgery-related operational data through a three-stage data comparison.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e1.3 Research Subjects and Data Sources\u003c/h2\u003e \u003cp\u003eThree types of data were selected as research samples: first, in vitro simulation data (including a needle insertion group and a tremor group, each divided into five subgroups. For the needle insertion group, standard needle insertion depths were set to 300 \u0026micro;m, 400 \u0026micro;m, 500 \u0026micro;m, 600 \u0026micro;m, and 700 \u0026micro;m; for the tremor group, standard tremor values were set to 50 \u0026micro;m, 100 \u0026micro;m, 150 \u0026micro;m, 200 \u0026micro;m, and 250 \u0026micro;m); second, ex vivo porcine eye experimental data (the needle insertion group was divided into five subgroups with dynamic tremor data, and standard needle insertion depths for the five subgroups were set to 300 \u0026micro;m, 400 \u0026micro;m, 500 \u0026micro;m, 600 \u0026micro;m, and 700 \u0026micro;m); third, ophthalmic surgical videos (containing dynamic processes related to needle insertion and tremor).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e1.4 Software Operation and Index Measurement\u003c/h2\u003e \u003cp\u003eThe aforementioned data were analyzed using the Fundus Surgery Image Evaluation Software and Kinovea software, respectively. The measured indices included needle insertion depth (\u0026micro;m), mean tremor value (\u0026micro;m), maximum tremor value (\u0026micro;m), minimum tremor value (\u0026micro;m), median tremor value (\u0026micro;m), and tremor variance (\u0026micro;m). All operations were performed by professional personnel to ensure the consistency of the measurement process.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e1.5.1 In Vitro Model\u003c/h2\u003e \u003cp\u003eThe in vitro model was used to measure the standard needle insertion depth and standard tremor value set for the instrument under a fixed blank background. The purpose of this model was to compare the accuracy of the measurement data from the two software programs in a simple background, excluding the interference of complex backgrounds and background movement. The in vitro data were divided into a needle insertion group and a tremor group, which were used to measure needle insertion and tremor, respectively; the tremor in the needle insertion group was only for reference and control. The needle insertion group was further divided into five subgroups, with standard needle insertion depths set to 300 \u0026micro;m, 400 \u0026micro;m, 500 \u0026micro;m, 600 \u0026micro;m, and 700 \u0026micro;m; each standard needle insertion depth was measured repeatedly for 5 times. The needle insertion depth was measured using the self-developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively. The tremor group was also divided into five subgroups, with standard tremor values set to 50 \u0026micro;m, 100 \u0026micro;m, 150 \u0026micro;m, 200 \u0026micro;m, and 250 \u0026micro;m; each standard tremor value was measured repeatedly for 5 times. The tremor values were measured using the self-developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively(Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e1.5.2 Ex - vivo Porcine Eye\u003c/h2\u003e \u003cp\u003e The ex vivo porcine eyes used in this study were by-products of food processing from qualified abattoirs, with sources complying with national relevant regulations and animal welfare requirements.\u003c/p\u003e \u003cp\u003eThe ex - vivo porcine eye was used to measure the standard needle insertion depth of the instrument into the fixed retina within the porcine eye and the tremor during the operation process. The aim was to compare the accuracy of the measurement data of the two groups of software in a background relatively close to the human retina, but without the interference of respiration and eye movement. The needle insertion group was divided into five subgroups, with standard needle insertion depths set to 300 \u0026micro;m, 400 \u0026micro;m, 500 \u0026micro;m, 600 \u0026micro;m, 700 \u0026micro;m, 800 \u0026micro;m, and 900 \u0026micro;m. Each standard needle insertion depth was measured repeatedly for 5 times, and the corresponding tremor value during each needle insertion was measured simultaneously. The measurements were performed using the self - developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively(Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e1.5.3 Clinical Surgical Videos\u003c/h2\u003e \u003cp\u003eTen ophthalmic surgical videos were selected, and the depth of instrument needle insertion into the retina and tremor were measured against the background of the patient\u0026rsquo;s moving retina. Measurements were performed using the self-developed Fundus Surgery Image Evaluation Software and Kinovea software, respectively. This group represents real clinical scenarios, with the living retina as the background\u0026mdash;characterized by complex and moving images. The software was required to accurately identify the instrument against the complex background and eliminate interferences such as respiratory movement or eye tremor(Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data Processing Results of the In Vitro Model\u003c/h2\u003e \u003cp\u003eThe experimental results of the needle insertion group are as follows (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e): In the standard depth segment of 300 \u0026micro;m (Serial Numbers 1\u0026ndash;5), the measured values of the self-developed Fundus Surgery Image Evaluation Software were concentrated in the range of 299.6\u0026ndash;305.6 \u0026micro;m, closely aligning with the blue dashed line. In contrast, the measured values of Kinovea were significantly higher (411.8\u0026ndash;495.6 \u0026micro;m), deviating from the standard line by approximately 40%\u0026ndash;65%. In the 400 \u0026micro;m segment (Serial Numbers 6\u0026ndash;10), the self-developed Fundus Surgery Image Evaluation Software continued to maintain measurements within the narrow range of 399.2\u0026ndash;402.0 \u0026micro;m, almost overlapping with the standard value; however, the results of Kinovea further increased to 583.9\u0026ndash;639.8 \u0026micro;m, with a deviation\u0026thinsp;\u0026gt;\u0026thinsp;45%. In the 500 \u0026micro;m segment (Serial Numbers 11\u0026ndash;15), the measured values of the self-developed Fundus Surgery Image Evaluation Software remained stable at 499.98\u0026ndash;501.7 \u0026micro;m, with an error\u0026thinsp;\u0026lt;\u0026thinsp;0.35%; in comparison, the results of Kinovea had increased to 734.6\u0026ndash;830.1 \u0026micro;m, showing a deviation of approximately 47%\u0026ndash;66%. In the 600 \u0026micro;m segment (Serial Numbers 16\u0026ndash;20), the self-developed Fundus Surgery Image Evaluation Software yielded readings of 600.0\u0026ndash;602.0 \u0026micro;m, maintaining consistency with the standard line; whereas Kinovea reached 903.5\u0026ndash;998.5 \u0026micro;m, with a deviation\u0026thinsp;\u0026gt;\u0026thinsp;50%. In the 700 \u0026micro;m segment (Serial Numbers 21\u0026ndash;25), the self-developed Fundus Surgery Image Evaluation Software measured 698.8\u0026ndash;703.6 \u0026micro;m, with an error\u0026thinsp;\u0026lt;\u0026thinsp;0.5%; the results of Kinovea had risen to 1068.9\u0026ndash;1173.4 \u0026micro;m, achieving a deviation of 50%\u0026ndash;68%.\u003c/p\u003e \u003cp\u003eIn summary, under the simple in vitro background, the measured needle insertion depths obtained by the self-developed Fundus Surgery Image Evaluation Software almost completely matched the set standards, demonstrating high accuracy. In contrast, Kinovea software exhibited systematic positive deviations under all standard depth conditions, and these deviations continued to expand as the set depth increased. This finding indicates that Kinovea still has significant range-related errors even in the absence of background interference.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eResults of the Tremor Group (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e): In the standard 50 \u0026micro;m segment (Serial Numbers 1\u0026ndash;5), the mean values measured by the self-developed Fundus Surgery Image Evaluation Software were stable at 46.9\u0026ndash;48.1 \u0026micro;m, closely adhering to the standard line. In contrast, the mean values of Kinovea were only 1.63\u0026ndash;3.05 \u0026micro;m, with a maximum value of merely 6.4\u0026ndash;18.0 \u0026micro;m, which was significantly lower than the standard value\u0026mdash;indicating that Kinovea could barely distinguish low-amplitude tremors. In the 100 \u0026micro;m segment (Serial Numbers 6\u0026ndash;10), the self-developed Fundus Surgery Image Evaluation Software yielded mean values of 86.5\u0026ndash;88.5 \u0026micro;m and maximum values of 96.3\u0026ndash;101.9 \u0026micro;m, still remaining close to the standard. However, Kinovea\u0026rsquo;s mean values were 1.63\u0026ndash;3.95 \u0026micro;m and maximum values were 5.56\u0026ndash;15.8 \u0026micro;m, which were still far below the standard, demonstrating that Kinovea still lacked sufficient sensitivity in this segment. In the 150 \u0026micro;m segment (Serial Numbers 11\u0026ndash;15), the self-developed Fundus Surgery Image Evaluation Software had mean values of 123.1\u0026ndash;130.1 \u0026micro;m and maximum values of 139\u0026ndash;144 \u0026micro;m, showing a slight underestimation but remaining identifiable. Kinovea, on the other hand, had mean values of 2.00\u0026ndash;4.09 \u0026micro;m and maximum values of 9.76\u0026ndash;17.1 \u0026micro;m, which were far below the standard and showed no significant difference from those in the 50 \u0026micro;m and 100 \u0026micro;m segments\u0026mdash;further confirming that Kinovea could not effectively capture low-to-moderate amplitude tremors. In the 200 \u0026micro;m segment (Serial Numbers 16\u0026ndash;20), the self-developed Fundus Surgery Image Evaluation Software produced mean values of 168.1\u0026ndash;174.2 \u0026micro;m and maximum values of 179\u0026ndash;190 \u0026micro;m, approaching the standard. Kinovea\u0026rsquo;s mean values were 2.11\u0026ndash;3.09 \u0026micro;m and maximum values were 8.37\u0026ndash;15.6 \u0026micro;m, still maintaining an extremely low level\u0026mdash;suggesting that Kinovea\u0026rsquo;s ability to identify high-frequency signals was continuously limited. In the 250 \u0026micro;m segment (Serial Numbers 21\u0026ndash;25), the self-developed Fundus Surgery Image Evaluation Software had mean values of 217.2\u0026ndash;218.2 \u0026micro;m and maximum values of 230\u0026ndash;236 \u0026micro;m, with a slight underestimation but a consistent trend. Kinovea\u0026rsquo;s mean values were 2.28\u0026ndash;3.29 \u0026micro;m and maximum values were 6.85\u0026ndash;14.7 \u0026micro;m, which were almost the same as those in the previous segments\u0026mdash;indicating that Kinovea could not effectively detect tremors throughout the entire measurement range.\u003c/p\u003e \u003cp\u003eIn summary, under the simple in vitro background, the self-developed Fundus Surgery Image Evaluation Software could effectively identify standard tremors ranging from 50 to 250 \u0026micro;m, with only a slight underestimation observed in the maximum amplitude segment. In contrast, the mean and maximum tremor values measured by Kinovea software in all standard segments were significantly lower than the standard values and remained at an extremely low level of \u0026lt;\u0026thinsp;5 \u0026micro;m. This indicates that Kinovea has insufficient resolution for tremor signals and cannot detect the actual tremor amplitude.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Data Processing Results of Ex Vivo Porcine Eyes\u003c/h2\u003e \u003cp\u003eThe measurement results of needle insertion depth in ex vivo porcine eyes are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e: For the 300 \u0026micro;m segment (Serial Numbers 1\u0026ndash;3), the self-developed Fundus Surgery Image Evaluation Software measured values ranging from 282.6 to 343.9 \u0026micro;m, which were slightly lower overall but fluctuated around the standard line; Kinovea measured values from 389.7 to 673.0 \u0026micro;m, significantly higher than the standard, with a maximum deviation\u0026thinsp;\u0026gt;\u0026thinsp;120%. For the 400 \u0026micro;m segment (Serial Numbers 4\u0026ndash;6), the self-developed Fundus Surgery Image Evaluation Software yielded readings of 409.9\u0026ndash;527.4 \u0026micro;m, which were basically consistent with the standard value; Kinovea gave readings of 633.1\u0026ndash;1020.8 \u0026micro;m, showing a systematic overestimation with a deviation\u0026thinsp;\u0026gt;\u0026thinsp;58%. For the 500 \u0026micro;m segment (Serial Numbers 7\u0026ndash;9), the self-developed Fundus Surgery Image Evaluation Software measured values of 510.1\u0026ndash;836.5 \u0026micro;m; although one measurement was relatively high (836.5 \u0026micro;m), it was still lower than the extreme value of 1287.1 \u0026micro;m measured by Kinovea; Kinovea generally overestimated by 55%\u0026ndash;157%. For the 600 \u0026micro;m segment (Serial Numbers 10\u0026ndash;12), the self-developed Fundus Surgery Image Evaluation Software obtained values of 505.5\u0026ndash;664.8 \u0026micro;m, close to the standard value; Kinovea measured 978.8\u0026ndash;1058.9 \u0026micro;m, overestimating by 63%\u0026ndash;76%. For the 700 \u0026micro;m segment (Serial Numbers 13\u0026ndash;15), the self-developed Fundus Surgery Image Evaluation Software recorded values of 647.4\u0026ndash;808.3 \u0026micro;m, slightly lower than or close to the standard; Kinovea continued to overestimate by 33%\u0026ndash;69% with values of 931.3\u0026ndash;1181.1 \u0026micro;m. For the 800 \u0026micro;m segment (Serial Numbers 16\u0026ndash;18), the self-developed Fundus Surgery Image Evaluation Software measured values of 773.1\u0026ndash;843.7 \u0026micro;m, fluctuating around the standard line; Kinovea showed one abnormally low value (373.1 \u0026micro;m), while the other values were still overestimated by 37%\u0026ndash;38%. For the 900 \u0026micro;m segment (Serial Numbers 19\u0026ndash;21), the self-developed Fundus Surgery Image Evaluation Software obtained values of 823.7\u0026ndash;1064 \u0026micro;m, slightly lower than or close to the standard; Kinovea had extremely large fluctuations with values of 236.1\u0026ndash;1684 \u0026micro;m, including two significant overestimations (1486 \u0026micro;m and 1684 \u0026micro;m).\u003c/p\u003e \u003cp\u003eIn summary, under the background of ex vivo porcine eyes, the measured needle insertion depths obtained by the self-developed Fundus Surgery Image Evaluation Software were generally consistent with the set standard values, and most deviations were controlled within \u0026plusmn;\u0026thinsp;20%. In contrast, Kinovea software exhibited systematic positive deviations and occasional abnormal underestimations. This indicates that both the measurement stability and accuracy of Kinovea in the background of complex tissues are lower than those of the self-developed Fundus Surgery Image Evaluation Software.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe tremor measurement results of ex vivo porcine eyes are shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e: For the 300 \u0026micro;m segment (Serial Numbers 1\u0026ndash;3): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 6\u0026ndash;11 \u0026micro;m and a maximum value of 32\u0026ndash;54 \u0026micro;m, both within the low-amplitude range; Kinovea had a mean value of 32\u0026ndash;55 \u0026micro;m and a maximum value of 60\u0026ndash;335 \u0026micro;m. The mean value of Kinovea was approximately 4\u0026ndash;8 times higher than that of the self-developed software, and an extreme peak of 335 \u0026micro;m appeared\u0026mdash;indicating that Kinovea misjudged background textures as high-amplitude tremors even in the segment with the lowest standard depth.\u003c/p\u003e \u003cp\u003eFor the 400 \u0026micro;m segment (Serial Numbers 4\u0026ndash;6): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 9\u0026ndash;26 \u0026micro;m and a maximum value of 60\u0026ndash;103 \u0026micro;m, showing a slight synchronous increase with the increase in depth; Kinovea had a mean value of 47\u0026ndash;103 \u0026micro;m and a maximum value of 59\u0026ndash;474 \u0026micro;m, with an extreme value of 474 \u0026micro;m, and the deviation further expanded to 3\u0026ndash;5 times.\u003c/p\u003e \u003cp\u003eFor the 500 \u0026micro;m segment (Serial Numbers 7\u0026ndash;9): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 9\u0026ndash;25 \u0026micro;m and a maximum value of 47\u0026ndash;103 \u0026micro;m, with little change in amplitude; Kinovea had a mean value of 0.7\u0026ndash;103 \u0026micro;m and a maximum value of 60\u0026ndash;159 \u0026micro;m, with severe fluctuations within the segment, indicating poor algorithm stability.\u003c/p\u003e \u003cp\u003eFor the 600 \u0026micro;m segment (Serial Numbers 10\u0026ndash;12): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 9\u0026ndash;13 \u0026micro;m and a maximum value of 46\u0026ndash;67 \u0026micro;m, maintaining low-amplitude stability; Kinovea had a mean value of 25\u0026ndash;68 \u0026micro;m and a maximum value of 29\u0026ndash;88 \u0026micro;m, continuing to be systematically 2\u0026ndash;3 times higher.\u003c/p\u003e \u003cp\u003eFor the 700 \u0026micro;m segment (Serial Numbers 13\u0026ndash;15): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 7\u0026ndash;12 \u0026micro;m and a maximum value of 46\u0026ndash;51 \u0026micro;m, with continued stable amplitude; Kinovea had a mean value of 46\u0026ndash;67 \u0026micro;m and a maximum value of 45\u0026ndash;67 \u0026micro;m. Although the gap with the self-developed software narrowed, it was still 3\u0026ndash;6 times higher overall.\u003c/p\u003e \u003cp\u003eFor the 800 \u0026micro;m segment (Serial Numbers 16\u0026ndash;18): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 5\u0026ndash;18 \u0026micro;m and a maximum value of 30\u0026ndash;155 \u0026micro;m, with a single-point extreme value of 155 \u0026micro;m; Kinovea had a mean value of 1\u0026ndash;155 \u0026micro;m and a maximum value of 23\u0026ndash;137 \u0026micro;m, with an abnormal low value of 1.7 \u0026micro;m and the largest fluctuations within the segment\u0026mdash;indicating that Kinovea had the worst stability when the depth further increased.\u003c/p\u003e \u003cp\u003eFor the 900 \u0026micro;m segment (Serial Numbers 19\u0026ndash;21): The self-developed Fundus Surgery Image Evaluation Software had a mean value of 10\u0026ndash;15 \u0026micro;m and a maximum value of 30\u0026ndash;61 \u0026micro;m, with amplitude returning to a low level; Kinovea had a mean value of 30\u0026ndash;50 \u0026micro;m and a maximum value of 30\u0026ndash;61 \u0026micro;m. Although there were no extreme peaks, it was still systematically 2\u0026ndash;3 times higher than the self-developed software.\u003c/p\u003e \u003cp\u003eIn summary, across the seven standard depth segments (300\u0026ndash;900 \u0026micro;m), the mean and maximum tremor values measured by the self-developed Fundus Surgery Image Evaluation Software remained\u0026thinsp;\u0026lt;\u0026thinsp;160 \u0026micro;m, with small differences between segments. This indicates that its algorithm can truly and stably capture the micro-tremor of the instrument against the complex retinal background. In contrast, Kinovea software, due to the sensitivity of its template matching algorithm to vascular textures and reflections, exhibited a systematic overestimation of 2\u0026ndash;8 times in all depth segments. It was also accompanied by extreme peaks of 300\u0026ndash;400 \u0026micro;m and abnormal low values of 1\u0026ndash;2 \u0026micro;m, which demonstrates that Kinovea has a severe deficiency in recognizing tremor signals against complex tissue backgrounds, resulting in low data reliability and inability to meet the requirements of high-precision retinal surgery evaluation.In the tremor comparison of ex vivo porcine eyes, the maximum tremor value of Kinovea for Sample No. 5 (corresponding to the 400 \u0026micro;m segment) was systematically recorded as 35,618.53 \u0026micro;m. This value was two orders of magnitude higher than the other data points in the same group and far exceeded the actual movement range of the instrument. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, this data point was disconnected from the overall curve, which can be clearly identified as an abnormal peak caused by Kinovea\u0026rsquo;s template matching algorithm misclassifying large-scale background drift or reflective flicker as \"tremor\". During subsequent statistical analysis, this outlier was excluded in accordance with the 3σ principle and no longer involved in the calculation of mean values and deviations.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Processing Results of Clinical Surgical Videos\u003c/h2\u003e \u003cp\u003eA total of ten clinical surgical videos were selected, and the measurement results of needle insertion depth after analyzing this group of videos are as follows (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDepth Range: The measurement results of the self-developed Fundus Surgery Image Evaluation Software ranged from 198 to 934 \u0026micro;m, all falling within the clinically common puncture range (\u0026lt;\u0026thinsp;1 mm); the measurement results of Kinovea ranged from 174 to 2107 \u0026micro;m, with a span of nearly 12 times, and the maximum value of 2107 \u0026micro;m (Serial Number 3) had exceeded the safe surgical limit.\u003c/p\u003e \u003cp\u003eDispersion and Outliers: The difference between the maximum and minimum measurements of the self-developed Fundus Surgery Image Evaluation Software was 736 \u0026micro;m, with no obvious outliers; the difference between the maximum and minimum measurements of Kinovea was 1932 \u0026micro;m, and an extremely high value of 2107 \u0026micro;m appeared at Serial Number 3, which could be determined as respiratory-ocular movement artifacts misjudged as needle insertion depth.\u003c/p\u003e \u003cp\u003eDeviation Trend: Among 8 cases (Serial Numbers 1, 2, 4, 5, 6, 7, 9, 10), the values of Kinovea were generally 50\u0026ndash;400 \u0026micro;m higher than those of the self-developed software, indicating a systematic positive deviation; in Serial Number 8, Kinovea showed a slightly lower value than the self-developed Fundus Surgery Image Evaluation Software (174 \u0026micro;m vs. 181 \u0026micro;m), which indicated occasional misalignment between the two algorithms.\u003c/p\u003e \u003cp\u003eIn summary, among the real surgical videos of manual operation group, the measurement results of the self-developed Fundus Surgery Image Evaluation Software were concentrated in the range of 200\u0026ndash;900 \u0026micro;m, which was consistent with clinical operation experience. The self-developed Fundus Surgery Image Evaluation Software could still stably output clinically acceptable depth data against the complex background of living organisms.In contrast, Kinovea could not effectively distinguish between the instrument tip and the moving retina, leading to a significant overestimation of depth. In particular, the abnormal value of 2107 \u0026micro;m at Serial Number 3 had exceeded the safety threshold, which was clinically unacceptable. It can be seen that Kinovea software was interfered by respiration and eye movement, resulting in systematic overestimation and extreme anomalies\u0026mdash;indicating that it is not suitable for direct application in the evaluation of in vivo retinal puncture depth.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe measurement results of needle insertion tremor in the manual operation group of clinical surgical videos are shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e, and the samples with similar performance can be grouped into three categories: low-amplitude stable group (No. 1, 4, 5, 6, 9, 10), medium-amplitude controllable group (No. 2, 7), and extreme abnormal group (No. 3, 8).For the low-amplitude stable group: The self-developed Fundus Surgery Image Evaluation Software measured a tremor mean value of 35\u0026ndash;80 \u0026micro;m and a maximum value of 114\u0026ndash;252 \u0026micro;m, all within the clinical safety range; Kinovea measured a tremor mean value of 111\u0026ndash;161 \u0026micro;m and a maximum value of 295\u0026ndash;682 \u0026micro;m, which was overall approximately 1.5\u0026ndash;4.6 times higher, showing consistent overestimation but no extreme peaks.For the medium-amplitude controllable group: The self-developed Fundus Surgery Image Evaluation Software measured a tremor mean value of 88\u0026ndash;98 \u0026micro;m and a maximum value of 277\u0026ndash;384 \u0026micro;m, still lower than the 400 \u0026micro;m warning line; Kinovea measured a tremor mean value of 152\u0026ndash;157 \u0026micro;m and a maximum value of 461\u0026ndash;507 \u0026micro;m, continuing to be 1.2\u0026ndash;1.8 times higher. Although the amplitude was larger, it did not exceed the safety upper limit.For the extreme abnormal group: The self-developed Fundus Surgery Image Evaluation Software measured a tremor mean value of 245 \u0026micro;m (No. 3) and 113 \u0026micro;m (No. 8), and a maximum value of 795 \u0026micro;m (No. 3) and 526 \u0026micro;m (No. 8), still within the acceptable range; in contrast, Kinovea measured a mean value of 577 \u0026micro;m and a maximum value of 2101 \u0026micro;m for No. 3, which had far exceeded the 1 mm safety limit; the maximum value measured for No. 8 by Kinovea was 433 \u0026micro;m, which was less extreme than No. 3 but still significantly higher than the corresponding value of the self-developed software.\u003c/p\u003e \u003cp\u003eIn summary, among the 10 cases of real surgical videos, 8 cases belong to the low-amplitude stable group and medium-amplitude controllable group, where Kinovea showed systematic overestimation with an amplitude of 1.2\u0026ndash;4.6 times. For the 2 cases in the extreme abnormal group, the maximum value of Kinovea for Case No. 3 exceeded 2 mm, which was clearly caused by misjudgment of respiratory-ocular movement drift and clinically unacceptable.The self-developed Fundus Surgery Image Evaluation Software had a mean value\u0026thinsp;\u0026lt;\u0026thinsp;250 \u0026micro;m and a maximum value\u0026thinsp;\u0026lt;\u0026thinsp;800 \u0026micro;m in all 10 cases, with concentrated and safe data\u0026mdash;demonstrating that it could reliably capture real tremors even against the complex background of living organisms. In contrast, Kinovea could not eliminate background movement, leading to widespread overestimation and occasional extreme anomalies, so it is not suitable for direct application in clinical tremor evaluation.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study found that in the video processing of in vitro model data, ex vivo porcine eye data, and clinical surgical videos, the values of needle insertion depth and tremor amplitude measured by the Fundus Surgery Image Evaluation Software were closer to the true values than those measured by Kinovea software. These results indicate that the Fundus Surgery Image Evaluation Software can more accurately assess the movement trajectory in delicate operations.Results from the three-stage tests all showed that in terms of needle insertion depth measurement, the data measured by the Fundus Surgery Image Evaluation Software was closer to the true value, while the values measured by Kinovea software were all larger than the set depth values. In terms of tremor amplitude measurement, the tremor values measured by both software were smaller than the set tremor values; however, compared with Kinovea software, the Fundus Surgery Image Evaluation Software was more capable of identifying the tremor amplitude and had better sensitivity.\u003c/p\u003e \u003cp\u003eThe Fundus Surgery Image Evaluation Software can automatically and more sensitively track the movement of surgical instruments, while Kinovea software cannot automatically identify the key points for tracking nor perform continuous detection. The tracking points of Kinovea software will deviate from the correct positions as the tracking time increases\u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e, thus requiring manual adjustment of the tracking points during long-term tracking\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e.The Fundus Surgery Image Evaluation Software can automatically identify key points and also support manual adjustment. After the software automatically identifies six points (such as the instrument tip), manual adjustment can be made if there is a deviation. The Fundus Surgery Image Evaluation Software uses the adjusted points as a reference for feedback, thereby achieving continuous self-optimization. This gives the software an advantage in processing video data of human surgical procedures.In ophthalmic surgery, the human eyeball shifts to a certain extent with the patient\u0026rsquo;s respiratory movement. Against this background, the software needs to track the surgical instruments moving within the vitreous cavity. Background movement will affect the software\u0026rsquo;s tracking of surgical instruments to a certain extent, leading to inaccurate measurement data. In previous studies, Kinovea software was mostly used in scenarios with a fixed background and the support of a high-definition camera; only under such conditions could Kinovea software maintain a certain level of stability. However, subtle background movements are unavoidable in ophthalmic surgery. Our Fundus Surgery Image Evaluation Software is designed specifically for ophthalmic surgery and has the advantage of handling simultaneous movements of both the surgical background and the surgical instruments.Additionally, the Fundus Surgery Image Evaluation Software achieves more accurate distance conversion. Kinovea software is mostly used in scenarios with camera shooting, while ophthalmic surgeries are performed under a microscope. There is a significant difference in lens distortion between ordinary cameras and microscopes; based on this, we have optimized the Fundus Surgery Image Evaluation Software.\u003c/p\u003e \u003cp\u003eWith the development of artificial intelligence technology, surgical robots are widely used in the medical field. For subretinal injection, robotic operation has obvious advantages; however, there has been no standard path as a reference for subretinal injection completed by robotic operation in previous studies.The self-developed Fundus Surgery Image Evaluation Software can conduct quantitative analysis on a large number of surgical videos from ophthalmology experts, thereby obtaining the optimal operation path and needle insertion depth for subretinal injection. Meanwhile, it can reduce operational tremor to a certain extent.In the future, the Fundus Surgery Image Evaluation Software can be applied to more surgical operations. By quantifying surgical videos, a relatively standard surgical operation mode can be obtained. Surgical robots can refer to this standard mode to independently complete some simple surgical steps.On the other hand, for surgical beginners, by learning the data analyzed by the Fundus Surgery Image Evaluation Software, they can learn more intuitively and master the key points of some surgical operations more quickly\u0026mdash;such as the operation path, needle insertion direction and angle, needle insertion depth, and injection speed of subretinal injection.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003eThe Fundus Surgery Image Evaluation Software is publicly accessible via oculotronics.net to support related research on ophthalmic microsurgery.\u003c/p\u003e\n\u003cp\u003eThe limitation of this study lies in that the current version of the Fundus Surgery Image Evaluation Software is only applicable to ophthalmic surgeries. It is adapted to microscopes used in ophthalmic procedures, and there may be certain errors when analyzing videos captured by other types of cameras.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003cbr\u003eAll authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers\u0026rsquo; bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study complied with the Declaration of Helsinki and was approved by the Ethics Committee of the Second Affiliated Hospital of Nanchang University. All participants provided written informed consent. The ex vivo porcine eyes used in this study were obtained from qualified and legitimate abattoirs as by-products of food processing. No animals were sacrificed specifically for this experiment, which complies with the requirements of animal welfare ethics. The experimental protocol has been approved by the Animal Ethics Committee of Nanchang University.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to publish\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed consent\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInformed consent was obtained from all individual participants included in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China(No.82360204) and the Jiangxi Provincial Natural Science Foundation General Program (No.20252BAC240512).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDel Amo EM, et al. Pharmacokinetic aspects of retinal drug delivery. Prog Retin Eye Res. 2017;57:134\u0026ndash;85.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng Y, Tang L, Zhou Y. Subretinal Injection: A Review on the Novel Route of Therapeutic Delivery for Vitreoretinal Diseases. Ophthalmic Res. 2017;58(4):217\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTripepi D et al. The Role of Subretinal Injection in Ophthalmic Surgery: Therapeutic Agent Delivery and Other Indications. Int J Mol Sci, 2023. 24(13).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIrigoyen C et al. Subretinal Injection Techniques for Retinal Disease: A Review. J Clin Med, 2022. 11(16).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSimunovic MP, et al. Two-step versus 1-step subretinal injection to compare subretinal drug delivery: a randomised study protocol. BMJ Open. 2021;11(12):e049976.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng J, et al. Subretinal injection of ranibizumab in advanced pediatric vasoproliferative disorders with total retinal detachments. Graefes Arch Clin Exp Ophthalmol. 2020;258(5):1005\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJeganathan VS, Shah S. Robotic technology in ophthalmic surgery. Curr Opin Ophthalmol. 2010;21(1):75\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe B, et al. A review of robotic surgical training: establishing a curriculum and credentialing process in ophthalmology. Eye (Lond). 2021;35(12):3192\u0026ndash;201.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoizenblatt M, et al. Robot-assisted tremor control for performance enhancement of retinal microsurgeons. Br J Ophthalmol. 2019;103(8):1195\u0026ndash;200.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaberley DAL, et al. A comparison of robotic and manual surgery for internal limiting membrane peeling. Graefes Arch Clin Exp Ophthalmol. 2020;258(4):773\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVorapojpisut S, et al. Quantifying sitting posture: A pilot feasibility study of computer vision and wearable sensors (Posture Lab) using a manikin model. Wearable Technol. 2025;6:e27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarcia-Ruiz P et al. Fiducial Objects: Custom Design and Evaluation. Sens (Basel), 2023. 23(24).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-ophthalmology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"boph","sideBox":"Learn more about [BMC Ophthalmology](http://bmcophthalmol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/boph","title":"BMC Ophthalmology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Fundus Surgery Image Evaluation Software, Kinovea, Needle Insertion Depth, Tremor Value, Ophthalmic Microsurgical Manipulation","lastPublishedDoi":"10.21203/rs.3.rs-8029636/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8029636/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eThis study systematically compares the performance of the self-developed \"Fundus Surgery Image Evaluation Software\" and the open-source tool Kinovea in quantifying key parameters of ophthalmic microsurgical operations.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThe self-developed \"Fundus Surgery Image Evaluation Software\" integrates modules for video cropping, key point annotation, reference frame verification, and automatic output, enabling simultaneous measurement of needle insertion depth and tremor amplitude. The performance of this software was evaluated through validation experiments (in vitro models, ex vivo porcine eyes, and clinical surgical videos) and compared with that of Kinovea. For the in vitro models, standard needle insertion depths of 300\u0026ndash;700 \u0026micro;m and standard tremor values of 50\u0026ndash;250 \u0026micro;m were set; ex vivo porcine eyes were used to simulate an environment similar to the human retinal environment, with measurements of needle insertion depth and tremor values within the same ranges; and clinical surgical videos were analyzed based on the dynamic retinal background in living subjects. The main measurement indicators included needle insertion depth, as well as the mean, maximum, minimum, median, and variance of tremor. The measurement deviation and stability of the two software tools were compared.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eIn all experimental scenarios, the measurement results of the self-developed \"Fundus Surgery Image Evaluation Software\" were significantly superior to those of Kinovea. In the in vitro scenario, the needle insertion error of the self-developed software was \u0026lt;\u0026thinsp;0.5%, while Kinovea exhibited a systematic positive bias of 40%\u0026ndash;68%. For the tremor segment, the self-developed software showed a slight underestimation but could identify tremors throughout the entire process, whereas Kinovea barely detected tremors (mean value\u0026thinsp;\u0026lt;\u0026thinsp;5 \u0026micro;m). In the ex vivo porcine eye experiment, the needle insertion deviation of the self-developed software was \u0026le;\u0026thinsp;\u0026plusmn;\u0026thinsp;20%, the mean tremor value was \u0026lt;\u0026thinsp;160 \u0026micro;m, and stability was maintained across segments. In contrast, Kinovea consistently overestimated the needle insertion depth and produced an extreme value of 1,500 \u0026micro;m; its tremor measurement was interfered with by background textures, leading to a 2\u0026ndash;8-fold overestimation and even an abnormal spike of 35,618 \u0026micro;m. In the clinical video analysis, the needle insertion depth measured by the self-developed software ranged from 198 to 934 \u0026micro;m, with a mean tremor value\u0026thinsp;\u0026lt;\u0026thinsp;250 \u0026micro;m and a maximum tremor value\u0026thinsp;\u0026lt;\u0026thinsp;800 \u0026micro;m, all falling within the safe range. For Kinovea, the maximum needle insertion depth reached 2,107 \u0026micro;m and the maximum tremor value was 2,101 \u0026micro;m; 8 out of 10 cases showed a systematic overestimation of 1.2\u0026ndash;4.6 folds.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThe self-developed \"Fundus Surgery Image Evaluation Software\" maintains high accuracy, low dispersion, and values within the clinically acceptable range under blank, tissue-based, and in vivo complex background conditions, making it safe for application in the quantitative evaluation of retinal microsurgery. Due to limitations in template matching, Kinovea exhibits significant systematic biases and extreme abnormalities, and thus is not suitable for direct use in the measurement of ophthalmic microsurgical procedures.\u003c/p\u003e","manuscriptTitle":"Analysis of Performance Differences Between Self-Developed Fundus Surgery Image Evaluation Software and Kinovea in Ophthalmic Microsurgery","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-22 08:47:51","doi":"10.21203/rs.3.rs-8029636/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-02-24T05:48:38+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-23T11:26:24+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-17T04:00:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"12191564716306901370301803805458851065","date":"2026-02-09T12:45:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"131431102294168549855508134992146426881","date":"2026-02-03T15:42:11+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-03T12:31:18+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"6127462866413436656496392190047061208","date":"2026-01-07T12:25:51+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-11T13:02:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-18T11:28:47+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-17T12:46:06+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Ophthalmology","date":"2025-11-17T12:41:40+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-ophthalmology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"boph","sideBox":"Learn more about [BMC Ophthalmology](http://bmcophthalmol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/boph","title":"BMC Ophthalmology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7f07e20f-8003-4786-b1af-0db2ab442bdc","owner":[],"postedDate":"December 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-02-24T05:55:32+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-22 08:47:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8029636","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8029636","identity":"rs-8029636","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}