AI-assisted Preoperative Planning Combined with Robotic-assisted Total Knee Arthroplasty vs. Conventional Surgery: A Retrospective Controlled Study

AI-assisted Preoperative Planning Combined with Robotic-assisted Total Knee Arthroplasty vs. Conventional Surgery: A Retrospective Controlled Study | 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 AI-assisted Preoperative Planning Combined with Robotic-assisted Total Knee Arthroplasty vs. Conventional Surgery: A Retrospective Controlled Study Yingdong Hu, Yuyu Fan, Zerui Sun, Hongxing Song This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6812822/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 16 Jan, 2026 Read the published version in BMC Musculoskeletal Disorders → Version 1 posted 14 You are reading this latest preprint version Abstract Objective: To compare perioperative outcomes and early functional recovery between AI-robotic and conventional total knee arthroplasty (TKA). Methods: We retrospectively analyzed data from 88 patients who underwent primary unilateral TKA for knee osteoarthritis between April 2024 and December 2024. The AI-robotic group (n=44) received AI-assisted preoperative planning and robot-assisted TKA, while the traditional group (n=44) underwent conventional 2D templating and manual TKA. Key metrics included preoperative prosthesis prediction accuracy, intraoperative and postoperative blood loss, osteotomy time, postoperative radiographic alignment, and functional scores. Results: The AI-robotic group showed significantly higher prosthesis prediction accuracy (femoral: 79.5% vs. 52.3%, P =0.023; tibial: 84.1% vs. 61.4%, P =0.042), shorter osteotomy time (15.24±4.71 vs. 18.43±4.76 minutes, P =0.002), reduced intraoperative blood loss (197.41±78.41 vs. 234.35±74.53 mL, P =0.026), and lower 72-hour total blood loss (1022.96±226.14 vs. 1118.71±193.30 mL, P =0.036). Postoperative lateral femoral component (LFC) angles were superior in the AI-robotic group (5.87±2.18° vs. 6.91±2.10°, P =0.025). At 6 weeks, the AI-robotic group had better VAS pain scores (2.27±1.12 vs. 2.84±1.22, P =0.029) and HSS scores (61.57±4.40 vs. 59.59±3.80, P =0.027). Conclusion: AI-assisted preoperative planning with robotic TKA improves prosthesis sizing accuracy, reduces perioperative blood loss and 72h total blood loss, and enhances early functional outcomes compared to conventional methods. These findings support AI-robotic integration as a precision solution for TKA, particularly in complex cases. Total knee replacement surgery Artificial Intelligence Robotic-assisted surgery Preoperative planning Figures Figure 1 Figure 2 Figure 3 1. Introduction Total knee arthroplasty (TKA) remains a highly effective treatment modality for end-stage knee osteoarthritis, with primary objectives to alleviate pain, correct deformities, and restore joint function [1] . The success of TKA hinges on three interdependent components: comprehensive preoperative planning, precise surgical execution, and systematic postoperative rehabilitation protocols. However, outcomes of conventional TKA procedures are disproportionately reliant on surgeon experience. Traditional preoperative planning methodologies, which involve manual measurement and analysis of two-dimensional (2D) X-ray films for prosthetic template selection and matching, inherently suffer from critical limitations. Specifically, 2D X-ray imaging is susceptible to magnification errors, projection angle discrepancies, and voltage fluctuations, all of which compromise its ability to provide accurate intraoperative guidance for prosthesis positioning. Such limitations may lead to intraoperative complications, including improper prosthesis sizing, osteotomy angle deviations, and soft tissue imbalance—factors directly linked to postoperative adverse events such as prosthesis loosening, joint instability, and abnormal wear. Additionally, significant intraoperative blood loss and subsequent postoperative bleeding risks frequently result in anemia, necessitate blood transfusions, and prolong convalescence. Enhancing preoperative planning precision, elevating surgical consistency, and optimizing patient outcomes therefore remain central challenges in the field of TKA. In recent years, with the profound integration of artificial intelligence (AI) technology into the orthopedic domain, its applications in preoperative planning and intraoperative navigation have offered novel solutions for enhancing the precision and personalization of total knee arthroplasty [2] . The AI KNEE software (Beijing Long Valley Medical Technology Co., Ltd., China), utilized in this study, employs artificial intelligence and three-dimensional (3D) image reconstruction technology to identify and segment key anatomical landmarks in patients' computed tomography (CT) images. The deep learning-based preoperative 3D imaging analysis system is capable of intelligently recommending prosthesis models by matching patients' anatomical parameters (e.g., femoral anteroposterior diameter, tibial plateau rotation angle). Intraoperatively, the real-time navigation technology provided by the ROPA KNEE robotic system (Beijing Long Valley Medical Technology Co., Ltd., China), adopted in this study, can dynamically optimize osteotomy trajectories to minimize bone loss and vascular injury risk, thereby reducing average intraoperative blood loss while shortening osteotomy duration. The clinical application of AI-assisted total knee arthroplasty (TKA) has been on the rise; however, limited research has systematically compared the benefits and drawbacks of AI-assisted preoperative planning combined with robotic-assisted TKA against conventional TKA. This retrospective study included 88 cases of primary unilateral TKA performed between April 2024 and December 2024, comprising 44 cases undergoing AI preoperative planning with robotic assistance and 44 cases receiving conventional surgery. The study compared perioperative parameters, postoperative imaging discrepancies, early functional outcomes, and patient satisfaction between the two groups, aiming to explore the comparative advantages and disadvantages of the AI-robotic TKA approach versus conventional procedures. 2. Materials and methods 2.1. Patient Selection Inclusion Criteria: (1) According to Kellgren-Lawrence X-ray grading criteria, the patient was diagnosed as end-stage osteoarthritis of the knee joint (Kellgren-Lawrence Grade IV) (2) Patients undergoing primary unilateral TKA. (3) Patients with complete data were followed up. Exclusion Criteria: (1) Severe comorbidities (e.g., coronary artery disease, severe hepatic or renal dysfunction). (2) Active infection in the knee or other sites. (3) Neuromuscular dysfunction (e.g., muscle atrophy, abductor weakness) or severe osteoporosis, metabolic bone disease, radiation-induced bone disease, or tumors around the hip joint. (4) Patients who have undergone knee joint surgery in the past. (5) Patients with incomplete follow-up data. A total of 88 consecutive eligible patients were enrolled in the study. Of these, 26 were male and 62 were female. All patients underwent bilateral weight-bearing knee radiographs and full-length anteroposterior lower limb radiographs, with a diagnosis of Kellgren-Lawrence grade IV knee osteoarthritis. Preoperatively, the benefits and risks of both surgical approaches were thoroughly explained to the patients and their families, who jointly selected the surgical method. Forty-four patients underwent AI-assisted preoperative planning combined with robotic-assisted total knee arthroplasty (AI-Robotic Group), while 44 patients received conventional manual surgery (Conventional Group). Complete imaging datasets, perioperative records, detailed preoperative planning documents, and postoperative follow-up data were available for all patients. All personally identifiable patient information was de-identified to ensure strict privacy protection, and all study procedures were conducted in full adherence to applicable laws and institutional regulations. This study was approved by the institutional review board (sjtkyll-lx-2022(87)) and performed in accordance with the ethical principles outlined in the 1989 version of the Declaration of Helsinki. 2.2. Preoperative Planning Methods All patients in the AI-Robotic group underwent preoperative full-length lower limb CT scans. Following de-identification of personal information, CT image data were uploaded to the AI KNEE software for preoperative planning. Artificial intelligence was employed to automatically segment osseous structures, reconstruct 3D models of the femur and tibia, identify bony anatomical landmarks of the knee joint, and automatically measure axes (e.g., tibial mechanical axis, femoral mechanical axis, tibial/femoral joint lines). Preoperative parameters were calculated, including femoral posterior condylar offset, valgus angle, posterior condylar angle, and tibial plateau posterior slope of the affected limb. The software generated anteroposterior and mediolateral diameters of the femoral condyle, automatically identified femoral condyle position/size, intelligently matched the most suitable femoral condyle prosthesis model from the TKA prosthesis database, and simulated prosthesis placement according to the calculated valgus angle. Additionally, the software intelligently analyzed tibial plateau morphology/size to automatically match the tibial plateau prosthesis, performed fine-tuning after simulated placement, automatically recognized osteotomy thickness based on prosthesis positioning, and selected the optimal insert matching the osteotomy thickness. Upon completion of planning, the system displayed the positioned femoral condyle prosthesis, tibial plateau component, and insert model, along with parameters including valgus angle, posterior condylar angle, tibial plateau posterior slope, osteotomy thickness, and femoral posterior condylar offset. (Fig. 1 ) The conventional group employed transparent two-dimensional (2D) prosthetic templates superimposed on radiographs for size selection, with the requirement that the femoral component correspond to the femoral condylar width within 1–2 mm in the coronal plane, anteroposterior dimensions within 1–2 mm in the sagittal plane, and curvature radius compatibility. Tibial component sizing adhered to analogous criteria for coronal alignment and anteroposterior fit. 2.3. Surgical Techniques For the AI-Robotic Group, patients were placed supine. After anesthesia, the surgical field was prepped and draped, and the ROPA KNEE system initialized. A midline knee incision with medial parapatellar arthrotomy exposed the joint. The ACL, PCL, menisci, and osteophytes were excised. Tibial/femoral trackers were affixed, and anatomical landmarks manually registered to coregister with preoperative CT for navigation validation. The robotic arm performed dynamic navigation-guided tibial osteotomy, followed by robot-assisted femoral cuts. Dynamic navigation aided posterior osteophyte resection and soft tissue releases. Trial components were inserted to assess gap balance, followed by canal preparation and trial implantation. Knee reduction evaluated alignment and ROM. Patellar tracking was optimized via lateral retinacular release, and osteophytes resected. Cemented components were implanted under pressure, with excess cement removed. Post-curing, ROM and alignment were rechecked. The wound was closed in layers, dressings applied, and tourniquet released with pulse verification before recovery. (Fig. 2 ) In the conventional group, patients underwent identical initial steps (positioning, anesthesia, sterilization, tourniquet application). A midline incision and medial parapatellar approach were used for joint exposure. ACL, PCL, menisci, and tibial osteophytes were excised. An extramedullary alignment rod was positioned along the tibial crest, with a posterior slope of 3°–7°, to perform proximal tibial osteotomy. Osteotomy surface flatness was assessed, and tibial component size was determined. Femoral condylar osteophytes were resected, followed by intramedullary canal entry anterior to the PCL origin and medial to the intercondylar notch. An intramedullary alignment rod set at 5°–7° valgus guided sequential distal femoral, anterior/posterior condylar, and chamfer cuts. Femoral component rotation was adjusted to 3° external rotation using the Insall line. Posterior femoral osteophytes and soft tissue releases were performed. Gap balancing, trial component installation, and ROM/alignment assessments mirrored the AI-robotic group. Cemented component fixation, patellar tracking optimization, and closure protocols were identical. 2.4 Primary Outcome Measures 2.4.1 Criteria for Judging the Accuracy of Preoperative Planning Prediction accuracy was defined as follows: (1) Accurate prediction: Preoperatively planned prosthesis size exactly matched the size used intraoperatively; (2) Moderate prediction: A discrepancy of ± 1 size between the preoperatively planned and intraoperatively used prosthesis; (3) Poor prediction: A discrepancy of ± 2 sizes between the preoperatively planned and intraoperatively used prosthesis. 2.4.2 Perioperative Data 2.4.2.1 General Patient Characteristics Demographic and clinical data were collected and recorded, including gender, age, body mass index (BMI), surgical side (left/right) and presence of comorbid chronic conditions (e.g., diabetes mellitus, hypertension). 2.4.2.2 Surgical Metrics Intraoperative blood loss, total operative time, osteotomy duration, and total blood loss within 72 hours postoperatively were recorded. The intraoperative blood loss of patients was measured by weighing method and calculated as follows. $$\:\text{T}\text{o}\text{t}\text{a}\text{l}\:\text{B}\text{l}\text{o}\text{o}\text{d}\:\text{L}\text{o}\text{s}\text{s}\:\left(\text{m}\text{L}\right)=\frac{\text{W}\text{e}\text{t}\:\text{W}\text{e}\text{i}\text{g}\text{h}\text{t}\:\text{o}\text{f}\:\text{B}\text{l}\text{o}\text{o}\text{d}-\text{S}\text{o}\text{a}\text{k}\text{e}\text{d}\:\text{I}\text{t}\text{e}\text{m}\text{s}\:\left(\text{g}\right)\:-\:\text{D}\text{r}\text{y}\:\text{W}\text{e}\text{i}\text{g}\text{h}\text{t}\:\left(\text{g}\right)}{1.05(\text{B}\text{l}\text{o}\text{o}\text{d}\:\text{D}\text{e}\text{n}\text{s}\text{i}\text{t}\text{y},\:\text{g}/\text{m}\text{L})}+\text{S}\text{u}\text{c}\text{t}\text{i}\text{o}\text{n}\:\text{V}\text{o}\text{l}\text{u}\text{m}\text{e}\:\left(\text{m}\text{L}\right)-\text{I}\text{r}\text{r}\text{i}\text{g}\text{a}\text{t}\text{i}\text{o}\text{n}\:\text{V}\text{o}\text{l}\text{u}\text{m}\text{e}\:\left(\text{m}\text{L}\right)\text{}$$ Preoperative and postoperative (72-hour) hematocrit (Hct) levels were recorded. Total blood loss was calculated using the Nadler equation [3] and Gross equation [4] . 2.4.2.3 Imaging-related Parameters (1) Hip-knee-ankle angle (HKA). Measured via weight-bearing X-ray of the lower limb, where H, K, and A represent the centers of the femoral head, knee joint, and ankle joint, respectively. The angle formed by the connection of these three points is the HKA, with a target angle of 180°. The difference between this angle and the target value is the HKA deviation angle. A deviation angle ≥ ± 3° indicates prosthetic misalignment. (2) Frontal femoral component angle (FFC), defined as the angle between the line connecting the medial and lateral condylar articular surfaces of the femoral prosthesis and the line connecting the knee joint center to the femoral head center. (3) Frontal tibial component angle (FTC), defined as the angle between the line connecting the medial and lateral articular surfaces of the tibial prosthesis and the tibial mechanical axis. (4) Lateral femoral component angle (LFC), defined as the angle between the femoral mechanical axis and the line connecting the medial and lateral condylar articular surfaces of the femoral prosthesis in the sagittal plane. (5) Lateral tibial component angle (LTC), also known as the posterior tilt angle, defined as the angle between the tibial mechanical axis and the articular surface of the tibial prosthesis in the sagittal plane. (Fig. 3 ) 2.4.3 Postoperative Follow-Up Data All patients underwent scheduled follow-ups at 6 weeks, 3 months, and 6 months postoperatively. Follow-up assessments focused on evaluating complications and functional recovery of the knee joint. Data collection combined in-person clinic visits and telephone interviews. Outcomes were systematically recorded using the Visual Analog Scale (VAS) for pain, knee range of motion (ROM), and the Hospital for Special Surgery (HSS) Knee Scoring System. 2.5 Statistical Analysis Data collection and analysis were performed using SPSS version 27.0 (IBM Corporation, SPSS, Armonk, NY, USA). All variables are expressed as the mean ± standard deviation or number and per-centage. Between-group differences in the mean values were compared using independent t-tests, and between group differences in the numbers and percentages were compared with chi-square tests or Fisher's exact test. P-values < 0.05 were considered statistically significant. 3. Results 3.1 Baseline Characteristics A total of 88 patients with end-stage knee osteoarthritis meeting the inclusion criteria were enrolled in this study. The cohort included 26 males (29.5%) and 62 females (70.5%), aged 58–80 years (mean age: 69.59 ± 4.65 years). Among them, 46 patients (52.3%) had comorbid chronic systemic conditions such as hypertension and diabetes. Surgical side distribution comprised 63 left knees (71.6%) and 25 right knees (28.4%), with all patients undergoing primary unilateral total knee arthroplasty. Preoperative imaging included bilateral standing lower extremity radiographs and full-length anteroposterior views, with all cases classified as Kellgren-Lawrence grade IV. Detailed baseline characteristics are summarized in Table 1 . Table 1 baseline of patients of two groups Variate AI-robotic group (44) Tradition group (44) P value Gender, n (%) 0.640 Male 14(31.8) 12(27.3) Female 30(68.2) 32(72.7) Age(years), mean ± SD 68.82 ± 5.18 70.36 ± 3.95 0.119 BMI, mean ± SD 26.06 ± 1.95 25.67 ± 1.84 0.334 Coexisting chronic diseases 0.200 Yes 20(45.5) 26(59.1) No 24(54.5) 18(40.9) Side of operation 0.478 Left 30(68.2) 33(75.0) Right 14(31.8) 11(25.0) BMI: Body mass index;SD༚Standard Deviation 3.2 Preoperative Prediction Accuracy Prediction accuracy was evaluated by comparing the prosthesis type selected intraoperatively with the preoperatively planned model. The AI-based planning group demonstrated higher prediction accuracy rates for both the femoral side (79.5%) and tibial side (84.1%). In contrast, the conventional 2D X-ray templating group showed lower accuracy rates of 52.3% for the femoral component and 61.4% for the tibial component. The differences in accuracy between the two groups were statistically significant ( P < 0.05). Detailed information are summarized in Table 2 . Table 2 Prediction accuracy Variate AI-robotic group (44) Traditional group (44) P value Prediction accuracy of the femoral side 0.023 Accurate 35(79.5) 23(52.3) General 8(18.2) 17(38.6) Poor 1(2.3) 4(9.1) Prediction accuracy of the tibial side 0.042 Accurate 37(84.1) 27(61.4) General 6(13.6) 14(31.8) Poor 1(2.3) 3(6.8) 3.3 Perioperative Indicators Perioperative outcomes are summarized in Table 3 . The mean operative time was 118.45 ± 27.21 minutes for the AI-robotic group and 112.86 ± 25.28 minutes for the traditional group, with no statistically significant difference between groups ( P > 0.05). However, the AI-robotic group demonstrated significantly shorter osteotomy time (15.24 ± 4.71 minutes vs. 18.43 ± 4.76 minutes, P = 0.002), reduced intraoperative blood loss (197.41 ± 78.41 mL vs. 234.35 ± 74.53 mL, P = 0.026), and lower total postoperative blood loss within 72 hours (1022.96 ± 226.14 mL vs. 1118.71 ± 193.30 mL, P = 0.036) compared to the traditional group. These findings indicate that AI-assisted planning combined with robotic surgery significantly improved intraoperative efficiency and reduced perioperative blood loss ( P < 0.05). Table 3 Peroperative indicators Variate AI-robotic group (44) Traditional group (44) P value Time of operation (min) 118.61 ± 27.21 112.86 ± 25.28 0.307 Time of osteotomy (min) 15.24 ± 4.71 18.43 ± 4.76 0.002 Peroperative bleeding (ml) 197.41 ± 78.41 234.35 ± 74.53 0.026 Total blood loss (ml) 1022.96 ± 226.14 1118.71 ± 193.30 0.036 3.4 Postoperative Radiographic Outcomes Detailed measurements are summarized in Table 4 . In the AI-robotic group, the postoperative HKA angle was 178.21 ± 1.10°, with a malalignment rate of 11.4%. The FFC, FTC, LFC, LTC angle were 84.25 ± 0.92°, 88.22 ± 0.88°, 5.87 ± 2.18°, and 84.71 ± 2.48°, respectively. Between-group comparisons demonstrated a statistically significant difference in the LFC angle ( P 0.05). Table 4 Postoperative imaging data Group Postoperative HKA( \(\:\stackrel{-}{x}\) ±s)/° HKA outlier (%) FFC( \(\:\stackrel{-}{x}\) ±s)/° FTC( \(\:\stackrel{-}{x}\) ±s)/° LFC( \(\:\stackrel{-}{x}\) ±s)/° LTC( \(\:\stackrel{-}{x}\) ±s)/° AI-robotic group (44) 178.21 ± 1.10 5(11.4) 84.25 ± 0.92 88.22 ± 0.88 5.87 ± 2.18 84.71 ± 2.48 Traditional group (44) 178.36 ± 0.91 6(13.6) 84.00 ± 1.11 88.16 ± 1.01 6.91 ± 2.10 85.03 ± 2.19 P value 0.488 0.747 0.803 0.690 0.025 0.530 3.5 Functional Follow-Up Outcomes Postoperative functional recovery outcomes are detailed in Table 5 . Comparative analysis of follow-up data between groups revealed that at 6 weeks postoperatively, the AI-robotic group exhibited significantly lower Visual Analog Scale (VAS) pain scores ( P < 0.05) and higher Hospital for Special Surgery (HSS) knee scores ( P < 0.05) compared to the traditional group. However, no statistically significant differences were observed in knee range of motion (ROM) between the two groups at 6 weeks, 3 months, or 6 months postoperatively ( P > 0.05). Additionally, both HSS and VAS scores showed no significant intergroup differences at 3 and 6 months postoperatively ( P > 0.05). Table 5 Postoperative functional recovery indicators Evaluation indicators Groups 6 weeks 3 months 6 months VAS AI-robotic group (44) 2.27 ± 1.12 1.18 ± 0.92 0.48 ± 0.55 Traditional group (44) 2.84 ± 1.22 1.55 ± 1.00 0.32 ± 0.56 P value 0.029 0.080 0.183 ROM (°) AI-robotic group (44) 106.03 ± 2.72 115.47 ± 4.20 122.71 ± 4.50 Traditional group (44) 105.07 ± 2.66 114.10 ± 4.12 121.41 ± 5.22 P value 0.100 0.125 0.215 HSS AI-robotic group (44) 61.57 ± 4.40 81.91 ± 2.73 92.09 ± 2.40 Traditional group (44) 59.59 ± 3.80 80.89 ± 2.98 91.16 ± 3.19 P value 0.027 0.097 0.125 4. Discussion Preoperative planning is a pivotal component in achieving precision-oriented TKA, as the quality of postoperative functional recovery directly hinges on the accuracy of prosthesis size matching and three-dimensional spatial alignment. Prosthesis size discrepancies can induce abnormal stress distribution, consequently leading to postoperative pain and functional impairment [5] . Research indicates that oversized prostheses carry risks of patellar overstuffing and excessive tension in bilateral collateral ligaments, whereas undersized prostheses compromise knee flexion stability and increase the likelihood of component loosening [6] . Although conventional X-ray templating remains widely used in clinical practice due to its operational simplicity, low technical threshold, and cost-effectiveness [7] , its limitations—including magnification distortion in two-dimensional radiographs, insufficient anatomical detail, lack of soft tissue-bony landmark visualization, and cumulative manual measurement errors—may result in significant discrepancies between preoperative plans and intraoperative anatomical realities [8, 9] . Arora et al. [10] demonstrated that preoperative planning with X-ray templating for TKA resulted in prosthesis size mismatches in 46.8% of cases, with interobserver and intraobserver variability rates of 46.8% and 43.6%, respectively. These findings highlight the observer-dependent and subjective nature of X-ray templating, which lacks both accuracy and reproducibility. Furthermore, traditional TKA preoperative planning is influenced not only by surgeon subjectivity but also by patient-specific variations [11] . Factors such as limb rotation/flexion during X-ray imaging, improper positioning, concomitant knee dysplasia, severe varus/valgus deformities, or joint instability often obscure bony landmarks and introduce measurement errors. In contrast, AI-based preoperative planning overcomes these limitations by generating precise 3D knee models from preoperative CT scans, enhancing visualization of anatomical landmarks and mitigating the drawbacks of conventional methods [12] . The AI system intelligently identifies anatomical landmarks, measures critical parameters (e.g., optimal osteotomy thickness), and matches prosthesis sizes from a digital database. Additionally, it simulates surgical outcomes in a 3D virtual environment preoperatively, enabling surgeons to refine plans, improve precision, reduce postoperative complications, extend prosthesis longevity, and enhance patient satisfaction [13] . Current studies have demonstrated the significant clinical utility of AI-assisted preoperative planning in total hip arthroplasty (THA) prosthesis sizing [14, 15] , but research on its application in TKA remains limited. The AI KNEE preoperative planning system used in this study integrates preoperative CT data with deep learning algorithms to optimize TKA planning. Results showed significantly higher accuracy rates for femoral (79.5% vs. 52.3%) and tibial (84.1% vs. 61.4%) component size prediction in the AI-assisted group compared with conventional 2D templating. Furthermore, the AI system provided 3D visualization and multi-parameter analysis (e.g., femoral valgus angle, tibial posterior slope), offering surgeons multidimensional decision support to minimize intraoperative uncertainties. While some studies suggest that a ± 1 size discrepancy is clinically acceptable for conventional templating—under which traditional X-ray methods achieve moderate match rates—this tolerance reflects an inherent compromise of legacy techniques. In contrast, AI-assisted systems enhance prosthesis matching precision, particularly in complex cases (e.g., severe bone defects, joint deformities), addressing the stringent anatomical adaptability requirements of precision and personalized medicine. Using a strict "zero-size deviation" criterion (± 0), our study confirmed that the AI system demonstrated superior accuracy with statistically significant advantages over conventional methods ( P < 0.05). These findings establish AI-assisted preoperative planning as a transformative approach for advancing surgical precision and technical innovation in TKA. Precise preoperative planning requires equally precise intraoperative execution to achieve optimal outcomes. The standardized, systematic framework of TKA procedures makes it particularly amenable to robotic-assisted technologies. Studies by Sires et al. [16] demonstrated that the Mako robotic system achieves high accuracy in TKA, with postoperative lower limb alignment deviations consistently within ± 3°. Similarly, Zhang et al. [17] reported that robotic-assisted TKA enables more accurate component positioning and improved intraoperative joint gap balancing compared with conventional techniques, leading to superior short-term patient outcomes. Additionally, robotic-assisted TKA holds advantages in enhancing surgical proficiency, managing operative pressure, and boosting surgeon confidence, thereby shortening the learning curve. Fu et al. [18] observed that patients undergoing robotic-assisted TKA exhibited higher postoperative HSS and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores, improved hip-knee-ankle (HKA) angles, and reduced operative times in follow-up exceeding six months. However, some studies note that robotic-assisted TKA may prolong operative times due to additional steps for robotic frame installation, registration protocols, and the initial learning curve [19–21] . In this study, data showed no significant difference in overall operative time between the robotic-assisted and conventional groups ( P = 0.307). However, the osteotomy phase was significantly shorter in the robotic group (15.24 ± 4.71 minutes vs. 18.43 ± 4.76 minutes, P = 0.002). This efficiency advantage may be attributed to four mechanisms: (1) Precision-Driven Osteotomy Minimizes Revisions: The robotic system enables high-accuracy osteotomy (error < 1°) through real-time bone volume monitoring (precision: 0.1 mm) and dynamic force feedback control, eliminating time-consuming intraoperative adjustments required in conventional techniques. (2) Seamless Preoperative-Intraoperative Data Integration: The 3D osteotomy plan generated by AI preoperative planning—including osteotomy volume, angles, and prosthesis sizing—was directly imported into the robotic navigation system, reducing redundant intraoperative anatomical landmark measurements. (3) Automated Osteotomy Path Execution: The robotic arm performed rapid bone resections along predefined trajectories, outperforming manual oscillating saws (which rely on surgeon experience). (4) Standardized Decision-Making Workflow: Augmented reality interfaces visually displayed osteotomy safety margins, simplifying surgical decision-making and minimizing delays from hesitation or misjudgment. Notably, while osteotomy efficiency improved, the additional time required for manual anatomical landmark registration and robotic system calibration partially offset these gains, resulting in comparable overall operative times between groups ( P > 0.05). Future advancements in human-robot interface design and registration algorithms are expected to further reduce total procedural duration. As a major orthopedic intervention, TKA is associated with substantial blood loss risks and subsequent transfusion-related complications, including immune reactions, infections, and disease transmission [22] , and studies report transfusion rates as high as 19% in TKA [23] , underscoring the critical need to mitigate bleeding risks in joint surgery. In this study, the robotic-assisted group demonstrated significantly lower intraoperative blood loss (197.41 ± 78.41 mL vs. 234.35 ± 74.53 mL, P = 0.026) and reduced total postoperative blood loss within 72 hours (1,022.96 ± 226.14 mL vs. 1,118.71 ± 193.30 mL, P = 0.036) compared with the conventional group, a hemostatic advantage attributed to the elimination of femoral intramedullary canal access—a mandatory step in traditional TKA for establishing the femoral mechanical axis but linked to medullary cavity trauma and heightened intraoperative bleeding—and to robotic-assisted TKA using mechanical arm navigation for osteotomy, which obviates medullary cavity penetration and reduces intraoperative hemorrhage. Additionally, the shorter osteotomy duration in the robotic group likely contributed to blood loss reduction by minimizing bone surface exposure time. While Khan et al . [24] proposed that robot-assisted TKA reduces intraoperative blood loss and transfusion rates via minimized soft tissue release and precise osteotomy, Held et al. [25] demonstrated significantly greater blood loss in robot-assisted TKA compared with conventional techniques when assessed via pre-/postoperative hematocrit and hemoglobin dynamics, highlighting the need for prospective studies with extended follow-up to validate the long-term hematoprotective effects of robotic technology and address interstudy discrepancies stemming from methodological variations in blood loss quantification and heterogeneous surgical protocols. In the imaging domain, this study revealed that patients undergoing TKA with AI-assisted preoperative planning and robotic navigation exhibited significantly more optimal LFC angles compared with the conventional group ( P < 0.05). The LFC angle, a critical indicator of osteotomy precision, reflects the positional relationship between the femoral prosthesis and the distal femoral anterior cortex. A smaller LFC angle may indicate intraoperative osteotomy notching, risking femoral fractures or prosthesis anterior tilt, whereas a larger angle may cause patellar overstuffing and postoperative pain [26, 27] . The AI-Robotic Group's improved LFC angle stemmed from three mechanisms: (1) AI planning using patient-specific CT to identify distal femoral anterior cortex notching points; (2) intraoperative registration verifying preoperative notching references to enhance accuracy; (3) robotic arm-assisted osteotomy with real-time volume monitoring for precise bone resection. Additionally, the AI-Robotic Group achieved postoperative hip-knee-ankle (HKA) angles closer to the ideal 180° mechanical alignment with a lower deviation rate, though statistical significance was not reached ( P > 0.05). The follow-up results of this study demonstrated that the AI-robotic group exhibited superior VAS pain scores (2.27 ± 1.12 vs. 2.84 ± 1.22, P = 0.029) and HSS scores (61.57 ± 4.40 vs. 59.59 ± 3.80, P = 0.027) compared to the conventional group at 6 weeks postoperatively, with this early functional advantage potentially stemming from multifactorial synergy: Firstly, the AI preoperative planning system precisely analyzed CT data to intelligently match the optimal prosthesis size while optimizing osteotomy surfaces, achieving anatomical alignment between the prosthesis geometry and the resected bone with higher surface coverage to reduce synovial inflammatory reactions caused by micromotion friction; second, the robotic system dynamically monitored flexion-extension gap pressure differences intraoperatively, adjusting osteotomy volume in real-time based on mechanical feedback to avoid excessive soft tissue release due to compensatory osteotomy errors common in traditional surgery; additionally, patient-specific joint line reconstruction based on 3D modeling accurately restored patellofemoral tracking, reducing contact pressure and consequently diminishing pain associated with abnormal mechanical loading; notably, precision osteotomy techniques also decreased bone debris release, potentially suppressing β-glycerophosphate phosphatase and proinflammatory cytokines (IL-6, TNF-α) expression levels to modulate the local inflammatory microenvironment and accelerate postoperative pain relief. However, the convergence of functional and pain scores between groups at 3 and 6 months postoperatively suggests that the core advantages of AI robotic technology may primarily focus on perioperative tissue protection and early biomechanical optimization, while long-term functional recovery depends more on patients' inherent rehabilitation capacity and soft tissue remodeling capabilities. Existing studies on robot-assisted TKA show partial consensus and differences, with Khlopas et al. [28] finding significantly better knee function (especially complex mobility) within 3 months postoperatively with Mako robot-assisted TKA than with conventional surgery, Xu et al.'s RCT [29] demonstrating superior tibial prosthesis alignment in the robotic group at 3 months postoperatively but noting that its correlation with functional scores or prosthesis survival rates still requires longer follow-up for validation, and long-term studies by Cho et al. [30] and Lee et al [31] . (average follow-up of 10–11 years) indicating that the robotic group had advantages in maintaining mechanical alignment and prosthesis positioning accuracy though differences in functional scores were not statistically significant, suggesting that imaging advantages may take longer to translate into improved clinical outcomes. The results of this study show better knee function scores within 6 weeks postoperatively with AI robot-assisted TKA but only confirm short-term benefits, with further extension of follow-up time needed to evaluate the mid- to long-term efficacy of robot-assisted TKA and determine whether the personalized alignment, reduced soft tissue release, and precision osteotomy achieved with AI robot assistance can translate into sustained postoperative outcomes. This study was a retrospective analysis. Although inherent selection bias risks existed, strict inclusion/exclusion criteria were prospectively defined preoperatively, and baseline data showed no significant statistical differences between groups in age, sex, BMI, and preoperative functional scores. All procedures were conducted by a single senior surgical team using a consistent prosthesis brand and model, with postoperative care delivered by a dedicated specialized nursing team following standardized protocols. These measures collectively mitigated selection bias risks to the fullest extent achievable. 5. Conclusion AI-assisted preoperative planning with robotic TKA improves prosthesis sizing accuracy, reduces perioperative blood loss and 72h total blood loss, and enhances early functional outcomes compared to conventional methods. These findings support AI-robotic integration as a precision solution for TKA, particularly in complex cases. Declarations Ethics approval and C onsent to participate This retrospective study was approved by the Ethics Committee of Beijing Shijitan Hospital, Capital Medical University (Ethics approval No: sjtkyll-lx-2022(87)) and conducted in accordance with the principles of the Declaration of Helsinki. Given the retrospective nature of the research and the use of de-identified patient data, the Ethics Committee waived the requirement for individual informed consent. Consent for publication Not applicable. Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Competing interests The authors declare that they have no competing interests. Funding No funding was received for this study. Authors' contributions Yingdong Hu: acquired data, analyzed and interpreted data, and wrote the manuscript. Hongxing Song: analyzed and interpreted data, and wrote the manuscript. Yuyu Fan: analyzed and interpreted data. Zerui Sun: acquired data. All authors read and approved the final manuscript. Acknowledgements Not applicable. Clinical trial number Not applicable. References Feng B, Zhu W, Bian YY, Chang X, Cheng KY, Weng XS: China artificial joint annual data report. Chin Med J (Engl) 2020, 134(6):752-753. von Eisenhart-Rothe R, Hinterwimmer F, Graichen H, Hirschmann MT: Artificial intelligence and robotics in TKA surgery: promising options for improved outcomes? Knee Surg Sports Traumatol Arthrosc 2022, 30(8):2535-2537. Nadler SB, Hidalgo JH, Bloch T: Prediction of blood volume in normal human adults. Surgery 1962, 51(2):224-232. Gross JB: Estimating allowable blood loss: corrected for dilution. Anesthesiology 1983, 58(3):277-280. Lan Q, Li S, Zhang J, Guo H, Yan L, Tang F: Reliable prediction of implant size and axial alignment in AI-based 3D preoperative planning for total knee arthroplasty. Sci Rep 2024, 14(1):16971. Tang A, Yeroushalmi D, Zak S, Lygrisse K, Schwarzkopf R, Meftah M: The effect of implant size difference on patient outcomes and failure after bilateral simultaneous total knee arthroplasty. J Orthop 2020, 22:282-287. Hsu AR, Gross CE, Bhatia S, Levine BR: Template-directed instrumentation in total knee arthroplasty: cost savings analysis. Orthopedics 2012, 35(11):e1596-1600. Conn KS, Clarke MT, Hallett JP: A simple guide to determine the magnification of radiographs and to improve the accuracy of preoperative templating. J Bone Joint Surg Br 2002, 84(2):269-272. Boese CK, Bredow J, Dargel J, Eysel P, Geiges H, Lechler P: Calibration Marker Position in Digital Templating of Total Hip Arthroplasty. J Arthroplasty 2016, 31(4):883-887. Arora J, Sharma S, Blyth M: The role of pre-operative templating in primary total knee replacement. Knee Surg Sports Traumatol Arthrosc 2005, 13(3):187-189. Goyal N, Stulberg SD: Evaluating the Precision of Preoperative Planning in Patient Specific Instrumentation: Can a Single MRI Yield Different Preoperative Plans? J Arthroplasty 2015, 30(7):1250-1253. Siddiqi A, Hardaker WM, Eachempati KK, Sheth NP: Advances in Computer-Aided Technology for Total Knee Arthroplasty. Orthopedics 2017, 40(6):338-352. Jones CW, Jerabek SA: Current Role of Computer Navigation in Total Knee Arthroplasty. J Arthroplasty 2018, 33(7):1989-1993. Chen X, Liu X, Wang Y, Ma R, Zhu S, Li S, Li S, Dong X, Li H, Wang G et al : Development and Validation of an Artificial Intelligence Preoperative Planning System for Total Hip Arthroplasty. Front Med (Lausanne) 2022, 9:841202. Huo J, Huang G, Han D, Wang X, Bu Y, Chen Y, Cai D, Zhao C: Value of 3D preoperative planning for primary total hip arthroplasty based on artificial intelligence technology. J Orthop Surg Res 2021, 16(1):156. Sires JD, Craik JD, Wilson CJ: Accuracy of Bone Resection in MAKO Total Knee Robotic-Assisted Surgery. J Knee Surg 2021, 34(7):745-748. Zhang J, Ndou WS, Ng N, Gaston P, Simpson PM, Macpherson GJ, Patton JT, Clement ND: Robotic-arm assisted total knee arthroplasty is associated with improved accuracy and patient reported outcomes: a systematic review and meta-analysis. Knee Surg Sports Traumatol Arthrosc 2022, 30(8):2677-2695. Fu X, She Y, Jin G, Liu C, Liu Z, Li W, Jin R: Comparison of robotic-assisted total knee arthroplasty: an updated systematic review and meta-analysis. J Robot Surg 2024, 18(1):292. Vermue H, Luyckx T, Winnock de Grave P, Ryckaert A, Cools AS, Himpe N, Victor J: Robot-assisted total knee arthroplasty is associated with a learning curve for surgical time but not for component alignment, limb alignment and gap balancing. Knee Surg Sports Traumatol Arthrosc 2022, 30(2):593-602. Naziri Q, Cusson BC, Chaudhri M, Shah NV, Sastry A: Making the transition from traditional to robotic-arm assisted TKA: What to expect? A single-surgeon comparative-analysis of the first-40 consecutive cases. J Orthop 2019, 16(4):364-368. Sodhi N, Khlopas A, Piuzzi NS, Sultan AA, Marchand RC, Malkani AL, Mont MA: The Learning Curve Associated with Robotic Total Knee Arthroplasty. J Knee Surg 2018, 31(1):17-21. Charoencholvanich K, Siriwattanasakul P: Tranexamic acid reduces blood loss and blood transfusion after TKA: a prospective randomized controlled trial. Clin Orthop Relat Res 2011, 469(10):2874-2880. Ma Q, Sun CJ, Wu S, Cai X: Comparison of Blood Loss between Open-Box Prosthesis and Closed-Box Prosthesis after Primary Total Knee Arthroplasty. Orthop Surg 2021, 13(3):768-777. Khan H, Dhillon K, Mahapatra P, Popat R, Zakieh O, Kim WJ, Nathwani D: Blood loss and transfusion risk in robotic-assisted knee arthroplasty: A retrospective analysis. Int J Med Robot 2021, 17(6):e2308. Held MB, Gazgalis A, Neuwirth AL, Shah RP, Cooper HJ, Geller JA: Imageless robotic-assisted total knee arthroplasty leads to similar 24-month WOMAC scores as compared to conventional total knee arthroplasty: a retrospective cohort study. Knee Surg Sports Traumatol Arthrosc 2022, 30(8):2631-2638. Jin Z, Wang Z, Xu K, Chu J, Xiang S, Tang Y, Wang R, Hua H, Zhang Z, Tong P et al : Effect of anterior femoral cortical notch grade on postoperative function and complications during TKA surgery: A multicenter, retrospective study. Open Med (Wars) 2024, 19(1):20240932. Okamoto Y, Nakajima M, Jotoku T, Otsuki S, Neo M: Capsular release around the intercondylar notch increases the extension gap in posterior-stabilized rotating-platform total knee arthroplasty. Knee 2016, 23(4):730-735. Khlopas A, Sodhi N, Hozack WJ, Chen AF, Mahoney OM, Kinsey T, Orozco F, Mont MA: Patient-Reported Functional and Satisfaction Outcomes after Robotic-Arm-Assisted Total Knee Arthroplasty: Early Results of a Prospective Multicenter Investigation. J Knee Surg 2020, 33(7):685-690. Xu J, Li L, Fu J, Xu C, Ni M, Chai W, Hao L, Zhang G, Chen J: Early Clinical and Radiographic Outcomes of Robot-Assisted Versus Conventional Manual Total Knee Arthroplasty: A Randomized Controlled Study. Orthop Surg 2022, 14(9):1972-1980. Cho KJ, Seon JK, Jang WY, Park CG, Song EK: Robotic versus conventional primary total knee arthroplasty: clinical and radiological long-term results with a minimum follow-up of ten years. Int Orthop 2019, 43(6):1345-1354. Lee YM, Kim GW, Lee CY, Song EK, Seon JK: No Difference in Clinical Outcomes and Survivorship for Robotic, Navigational, and Conventional Primary Total Knee Arthroplasty with a Minimum Follow-up of 10 Years. Clin Orthop Surg 2023, 15(1):82-91. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 16 Jan, 2026 Read the published version in BMC Musculoskeletal Disorders → Version 1 posted Editorial decision: Revision requested 22 Jul, 2025 Reviews received at journal 19 Jul, 2025 Reviewers agreed at journal 13 Jul, 2025 Reviews received at journal 13 Jul, 2025 Reviewers agreed at journal 13 Jul, 2025 Reviews received at journal 07 Jul, 2025 Reviewers agreed at journal 23 Jun, 2025 Reviewers agreed at journal 22 Jun, 2025 Reviewers agreed at journal 15 Jun, 2025 Reviewers invited by journal 13 Jun, 2025 Editor invited by journal 05 Jun, 2025 Editor assigned by journal 05 Jun, 2025 Submission checks completed at journal 05 Jun, 2025 First submitted to journal 03 Jun, 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-6812822","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":471591254,"identity":"0af5d7a2-cc31-43fc-b525-de5086eaf232","order_by":0,"name":"Yingdong Hu","email":"","orcid":"","institution":"Capital Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yingdong","middleName":"","lastName":"Hu","suffix":""},{"id":471591255,"identity":"a3d792fb-a36a-427a-ad1f-bf707480957a","order_by":1,"name":"Yuyu Fan","email":"","orcid":"","institution":"Capital Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yuyu","middleName":"","lastName":"Fan","suffix":""},{"id":471591256,"identity":"db63a67d-ae96-4338-88db-3ebe1e5e3bd6","order_by":2,"name":"Zerui Sun","email":"","orcid":"","institution":"Capital Medical University","correspondingAuthor":false,"prefix":"","firstName":"Zerui","middleName":"","lastName":"Sun","suffix":""},{"id":471591257,"identity":"90775601-628e-4fcf-a1ab-b1f5afa08e3a","order_by":3,"name":"Hongxing Song","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIie3QsarCMBSA4VMCR4ejbhJwyCtECiJYfBYvQqcO181dqEvdK4K+go9QOeAUnYW7CB1dAi6OejcXbUbBfFvg/CQ5AJ73oQSAHGJtVhT25p5E4ybtf3bLzD2Jg7VMQq6jw7iaL/gySVkgJJaBQLXaxftEm2M8yA03EQ5b/u1Dd7kaVSQy6YU0fdwSLLacE4z0X0Wi8v9Ec5AKOjOhQwKnJCxpGgcpErgl2pieIBONkVA/liyr/6LmWXhtpHKoNmVp7S1SrU7VwwBQPh3ky7FnwjqNeZ7nfa87F4VB+iO8YMgAAAAASUVORK5CYII=","orcid":"","institution":"Capital Medical University","correspondingAuthor":true,"prefix":"","firstName":"Hongxing","middleName":"","lastName":"Song","suffix":""}],"badges":[],"createdAt":"2025-06-03 15:23:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6812822/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6812822/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12891-025-09455-5","type":"published","date":"2026-01-16T16:30:59+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":84919164,"identity":"645caeca-db77-4623-92b4-5ea739415bd6","added_by":"auto","created_at":"2025-06-18 19:32:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":121486,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAI -assisted preoperative planning\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6812822/v1/ff5d8899d139b568a1381cf3.png"},{"id":84920945,"identity":"55c142fe-5092-449f-903c-257be6ba063e","added_by":"auto","created_at":"2025-06-18 19:40:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":83538,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRobotic arm-assisted navigation-guided osteotomy\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6812822/v1/87584b7f5d831f3af09552eb.png"},{"id":84919165,"identity":"abd666b4-6c29-4edc-a149-7a912239e54d","added_by":"auto","created_at":"2025-06-18 19:32:41","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":203369,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImaging index measurement diagram a. Hip knee ankle angle; b. Later femoral component（LFC）and later tibial component（LTC）on sagittal view; c. Frontal femoral component（FFC）on coronal view; d. Frontaltibal component（FTC）on coronal view.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6812822/v1/6235ca037210d5f94fe17540.png"},{"id":100614913,"identity":"8c476e01-acef-4c11-9852-2acb0cab00d7","added_by":"auto","created_at":"2026-01-19 17:28:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1626782,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6812822/v1/56959bb6-8643-4162-b082-323121d6a685.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"AI-assisted Preoperative Planning Combined with Robotic-assisted Total Knee Arthroplasty vs. Conventional Surgery: A Retrospective Controlled Study","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eTotal knee arthroplasty (TKA) remains a highly effective treatment modality for end-stage knee osteoarthritis, with primary objectives to alleviate pain, correct deformities, and restore joint function \u003csup\u003e[1]\u003c/sup\u003e. The success of TKA hinges on three interdependent components: comprehensive preoperative planning, precise surgical execution, and systematic postoperative rehabilitation protocols. However, outcomes of conventional TKA procedures are disproportionately reliant on surgeon experience. Traditional preoperative planning methodologies, which involve manual measurement and analysis of two-dimensional (2D) X-ray films for prosthetic template selection and matching, inherently suffer from critical limitations. Specifically, 2D X-ray imaging is susceptible to magnification errors, projection angle discrepancies, and voltage fluctuations, all of which compromise its ability to provide accurate intraoperative guidance for prosthesis positioning. Such limitations may lead to intraoperative complications, including improper prosthesis sizing, osteotomy angle deviations, and soft tissue imbalance\u0026mdash;factors directly linked to postoperative adverse events such as prosthesis loosening, joint instability, and abnormal wear. Additionally, significant intraoperative blood loss and subsequent postoperative bleeding risks frequently result in anemia, necessitate blood transfusions, and prolong convalescence. Enhancing preoperative planning precision, elevating surgical consistency, and optimizing patient outcomes therefore remain central challenges in the field of TKA.\u003c/p\u003e \u003cp\u003eIn recent years, with the profound integration of artificial intelligence (AI) technology into the orthopedic domain, its applications in preoperative planning and intraoperative navigation have offered novel solutions for enhancing the precision and personalization of total knee arthroplasty \u003csup\u003e[2]\u003c/sup\u003e. The AI KNEE software (Beijing Long Valley Medical Technology Co., Ltd., China), utilized in this study, employs artificial intelligence and three-dimensional (3D) image reconstruction technology to identify and segment key anatomical landmarks in patients' computed tomography (CT) images. The deep learning-based preoperative 3D imaging analysis system is capable of intelligently recommending prosthesis models by matching patients' anatomical parameters (e.g., femoral anteroposterior diameter, tibial plateau rotation angle). Intraoperatively, the real-time navigation technology provided by the ROPA KNEE robotic system (Beijing Long Valley Medical Technology Co., Ltd., China), adopted in this study, can dynamically optimize osteotomy trajectories to minimize bone loss and vascular injury risk, thereby reducing average intraoperative blood loss while shortening osteotomy duration.\u003c/p\u003e \u003cp\u003eThe clinical application of AI-assisted total knee arthroplasty (TKA) has been on the rise; however, limited research has systematically compared the benefits and drawbacks of AI-assisted preoperative planning combined with robotic-assisted TKA against conventional TKA. This retrospective study included 88 cases of primary unilateral TKA performed between April 2024 and December 2024, comprising 44 cases undergoing AI preoperative planning with robotic assistance and 44 cases receiving conventional surgery. The study compared perioperative parameters, postoperative imaging discrepancies, early functional outcomes, and patient satisfaction between the two groups, aiming to explore the comparative advantages and disadvantages of the AI-robotic TKA approach versus conventional procedures.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Patient Selection\u003c/h2\u003e \u003cp\u003eInclusion Criteria: (1) According to Kellgren-Lawrence X-ray grading criteria, the patient was diagnosed as end-stage osteoarthritis of the knee joint (Kellgren-Lawrence Grade IV) (2) Patients undergoing primary unilateral TKA. (3) Patients with complete data were followed up.\u003c/p\u003e \u003cp\u003eExclusion Criteria: (1) Severe comorbidities (e.g., coronary artery disease, severe hepatic or renal dysfunction). (2) Active infection in the knee or other sites. (3) Neuromuscular dysfunction (e.g., muscle atrophy, abductor weakness) or severe osteoporosis, metabolic bone disease, radiation-induced bone disease, or tumors around the hip joint. (4) Patients who have undergone knee joint surgery in the past. (5) Patients with incomplete follow-up data.\u003c/p\u003e \u003cp\u003eA total of 88 consecutive eligible patients were enrolled in the study. Of these, 26 were male and 62 were female. All patients underwent bilateral weight-bearing knee radiographs and full-length anteroposterior lower limb radiographs, with a diagnosis of Kellgren-Lawrence grade IV knee osteoarthritis. Preoperatively, the benefits and risks of both surgical approaches were thoroughly explained to the patients and their families, who jointly selected the surgical method. Forty-four patients underwent AI-assisted preoperative planning combined with robotic-assisted total knee arthroplasty (AI-Robotic Group), while 44 patients received conventional manual surgery (Conventional Group). Complete imaging datasets, perioperative records, detailed preoperative planning documents, and postoperative follow-up data were available for all patients. All personally identifiable patient information was de-identified to ensure strict privacy protection, and all study procedures were conducted in full adherence to applicable laws and institutional regulations. This study was approved by the institutional review board (sjtkyll-lx-2022(87)) and performed in accordance with the ethical principles outlined in the 1989 version of the Declaration of Helsinki.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Preoperative Planning Methods\u003c/h2\u003e \u003cp\u003eAll patients in the AI-Robotic group underwent preoperative full-length lower limb CT scans. Following de-identification of personal information, CT image data were uploaded to the AI KNEE software for preoperative planning. Artificial intelligence was employed to automatically segment osseous structures, reconstruct 3D models of the femur and tibia, identify bony anatomical landmarks of the knee joint, and automatically measure axes (e.g., tibial mechanical axis, femoral mechanical axis, tibial/femoral joint lines). Preoperative parameters were calculated, including femoral posterior condylar offset, valgus angle, posterior condylar angle, and tibial plateau posterior slope of the affected limb. The software generated anteroposterior and mediolateral diameters of the femoral condyle, automatically identified femoral condyle position/size, intelligently matched the most suitable femoral condyle prosthesis model from the TKA prosthesis database, and simulated prosthesis placement according to the calculated valgus angle. Additionally, the software intelligently analyzed tibial plateau morphology/size to automatically match the tibial plateau prosthesis, performed fine-tuning after simulated placement, automatically recognized osteotomy thickness based on prosthesis positioning, and selected the optimal insert matching the osteotomy thickness. Upon completion of planning, the system displayed the positioned femoral condyle prosthesis, tibial plateau component, and insert model, along with parameters including valgus angle, posterior condylar angle, tibial plateau posterior slope, osteotomy thickness, and femoral posterior condylar offset. (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe conventional group employed transparent two-dimensional (2D) prosthetic templates superimposed on radiographs for size selection, with the requirement that the femoral component correspond to the femoral condylar width within 1\u0026ndash;2 mm in the coronal plane, anteroposterior dimensions within 1\u0026ndash;2 mm in the sagittal plane, and curvature radius compatibility. Tibial component sizing adhered to analogous criteria for coronal alignment and anteroposterior fit.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Surgical Techniques\u003c/h2\u003e \u003cp\u003eFor the AI-Robotic Group, patients were placed supine. After anesthesia, the surgical field was prepped and draped, and the ROPA KNEE system initialized. A midline knee incision with medial parapatellar arthrotomy exposed the joint. The ACL, PCL, menisci, and osteophytes were excised. Tibial/femoral trackers were affixed, and anatomical landmarks manually registered to coregister with preoperative CT for navigation validation. The robotic arm performed dynamic navigation-guided tibial osteotomy, followed by robot-assisted femoral cuts. Dynamic navigation aided posterior osteophyte resection and soft tissue releases. Trial components were inserted to assess gap balance, followed by canal preparation and trial implantation. Knee reduction evaluated alignment and ROM. Patellar tracking was optimized via lateral retinacular release, and osteophytes resected. Cemented components were implanted under pressure, with excess cement removed. Post-curing, ROM and alignment were rechecked. The wound was closed in layers, dressings applied, and tourniquet released with pulse verification before recovery. (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the conventional group, patients underwent identical initial steps (positioning, anesthesia, sterilization, tourniquet application). A midline incision and medial parapatellar approach were used for joint exposure. ACL, PCL, menisci, and tibial osteophytes were excised. An extramedullary alignment rod was positioned along the tibial crest, with a posterior slope of 3\u0026deg;\u0026ndash;7\u0026deg;, to perform proximal tibial osteotomy. Osteotomy surface flatness was assessed, and tibial component size was determined. Femoral condylar osteophytes were resected, followed by intramedullary canal entry anterior to the PCL origin and medial to the intercondylar notch. An intramedullary alignment rod set at 5\u0026deg;\u0026ndash;7\u0026deg; valgus guided sequential distal femoral, anterior/posterior condylar, and chamfer cuts. Femoral component rotation was adjusted to 3\u0026deg; external rotation using the Insall line. Posterior femoral osteophytes and soft tissue releases were performed. Gap balancing, trial component installation, and ROM/alignment assessments mirrored the AI-robotic group. Cemented component fixation, patellar tracking optimization, and closure protocols were identical.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Primary Outcome Measures\u003c/h2\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e2.4.1 Criteria for Judging the Accuracy of Preoperative Planning\u003c/h2\u003e \u003cp\u003ePrediction accuracy was defined as follows: (1) Accurate prediction: Preoperatively planned prosthesis size exactly matched the size used intraoperatively; (2) Moderate prediction: A discrepancy of \u0026plusmn;\u0026thinsp;1 size between the preoperatively planned and intraoperatively used prosthesis; (3) Poor prediction: A discrepancy of \u0026plusmn;\u0026thinsp;2 sizes between the preoperatively planned and intraoperatively used prosthesis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e2.4.2 Perioperative Data\u003c/h2\u003e \u003cdiv id=\"Sec9\" class=\"Section4\"\u003e \u003ch2\u003e2.4.2.1 General Patient Characteristics\u003c/h2\u003e \u003cp\u003eDemographic and clinical data were collected and recorded, including gender, age, body mass index (BMI), surgical side (left/right) and presence of comorbid chronic conditions (e.g., diabetes mellitus, hypertension).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section4\"\u003e \u003ch2\u003e2.4.2.2 Surgical Metrics\u003c/h2\u003e \u003cp\u003eIntraoperative blood loss, total operative time, osteotomy duration, and total blood loss within 72 hours postoperatively were recorded.\u003c/p\u003e \u003cp\u003eThe intraoperative blood loss of patients was measured by weighing method and calculated as follows.\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{T}\\text{o}\\text{t}\\text{a}\\text{l}\\:\\text{B}\\text{l}\\text{o}\\text{o}\\text{d}\\:\\text{L}\\text{o}\\text{s}\\text{s}\\:\\left(\\text{m}\\text{L}\\right)=\\frac{\\text{W}\\text{e}\\text{t}\\:\\text{W}\\text{e}\\text{i}\\text{g}\\text{h}\\text{t}\\:\\text{o}\\text{f}\\:\\text{B}\\text{l}\\text{o}\\text{o}\\text{d}-\\text{S}\\text{o}\\text{a}\\text{k}\\text{e}\\text{d}\\:\\text{I}\\text{t}\\text{e}\\text{m}\\text{s}\\:\\left(\\text{g}\\right)\\:-\\:\\text{D}\\text{r}\\text{y}\\:\\text{W}\\text{e}\\text{i}\\text{g}\\text{h}\\text{t}\\:\\left(\\text{g}\\right)}{1.05(\\text{B}\\text{l}\\text{o}\\text{o}\\text{d}\\:\\text{D}\\text{e}\\text{n}\\text{s}\\text{i}\\text{t}\\text{y},\\:\\text{g}/\\text{m}\\text{L})}+\\text{S}\\text{u}\\text{c}\\text{t}\\text{i}\\text{o}\\text{n}\\:\\text{V}\\text{o}\\text{l}\\text{u}\\text{m}\\text{e}\\:\\left(\\text{m}\\text{L}\\right)-\\text{I}\\text{r}\\text{r}\\text{i}\\text{g}\\text{a}\\text{t}\\text{i}\\text{o}\\text{n}\\:\\text{V}\\text{o}\\text{l}\\text{u}\\text{m}\\text{e}\\:\\left(\\text{m}\\text{L}\\right)\\text{}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ePreoperative and postoperative (72-hour) hematocrit (Hct) levels were recorded. Total blood loss was calculated using the Nadler equation \u003csup\u003e[3]\u003c/sup\u003e and Gross equation \u003csup\u003e[4]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section4\"\u003e \u003ch2\u003e2.4.2.3 Imaging-related Parameters\u003c/h2\u003e \u003cp\u003e(1) Hip-knee-ankle angle (HKA). Measured via weight-bearing X-ray of the lower limb, where H, K, and A represent the centers of the femoral head, knee joint, and ankle joint, respectively. The angle formed by the connection of these three points is the HKA, with a target angle of 180\u0026deg;. The difference between this angle and the target value is the HKA deviation angle. A deviation angle\u0026thinsp;\u0026ge;\u0026thinsp;\u0026plusmn;\u0026thinsp;3\u0026deg; indicates prosthetic misalignment. (2) Frontal femoral component angle (FFC), defined as the angle between the line connecting the medial and lateral condylar articular surfaces of the femoral prosthesis and the line connecting the knee joint center to the femoral head center. (3) Frontal tibial component angle (FTC), defined as the angle between the line connecting the medial and lateral articular surfaces of the tibial prosthesis and the tibial mechanical axis. (4) Lateral femoral component angle (LFC), defined as the angle between the femoral mechanical axis and the line connecting the medial and lateral condylar articular surfaces of the femoral prosthesis in the sagittal plane. (5) Lateral tibial component angle (LTC), also known as the posterior tilt angle, defined as the angle between the tibial mechanical axis and the articular surface of the tibial prosthesis in the sagittal plane. (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 \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e2.4.3 Postoperative Follow-Up Data\u003c/h2\u003e \u003cp\u003eAll patients underwent scheduled follow-ups at 6 weeks, 3 months, and 6 months postoperatively. Follow-up assessments focused on evaluating complications and functional recovery of the knee joint. Data collection combined in-person clinic visits and telephone interviews. Outcomes were systematically recorded using the Visual Analog Scale (VAS) for pain, knee range of motion (ROM), and the Hospital for Special Surgery (HSS) Knee Scoring System.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Statistical Analysis\u003c/h2\u003e \u003cp\u003eData collection and analysis were performed using SPSS version 27.0 (IBM Corporation, SPSS, Armonk, NY, USA). All variables are expressed as the mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation or number and per-centage. Between-group differences in the mean values were compared using independent t-tests, and between group differences in the numbers and percentages were compared with chi-square tests or Fisher's exact test. P-values\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were considered statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Baseline Characteristics\u003c/h2\u003e \u003cp\u003eA total of 88 patients with end-stage knee osteoarthritis meeting the inclusion criteria were enrolled in this study. The cohort included 26 males (29.5%) and 62 females (70.5%), aged 58\u0026ndash;80 years (mean age: 69.59\u0026thinsp;\u0026plusmn;\u0026thinsp;4.65 years). Among them, 46 patients (52.3%) had comorbid chronic systemic conditions such as hypertension and diabetes. Surgical side distribution comprised 63 left knees (71.6%) and 25 right knees (28.4%), with all patients undergoing primary unilateral total knee arthroplasty. Preoperative imaging included bilateral standing lower extremity radiographs and full-length anteroposterior views, with all cases classified as Kellgren-Lawrence grade IV. Detailed baseline characteristics are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ebaseline of patients of two groups\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTradition group (44)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.640\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14(31.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12(27.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30(68.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e32(72.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge(years), mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e68.82\u0026thinsp;\u0026plusmn;\u0026thinsp;5.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e70.36\u0026thinsp;\u0026plusmn;\u0026thinsp;3.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.119\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e26.06\u0026thinsp;\u0026plusmn;\u0026thinsp;1.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e25.67\u0026thinsp;\u0026plusmn;\u0026thinsp;1.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.334\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCoexisting chronic diseases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.200\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e20(45.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e26(59.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e24(54.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e18(40.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSide of operation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.478\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30(68.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e33(75.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14(31.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11(25.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eBMI: Body mass index;SD༚Standard Deviation\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Preoperative Prediction Accuracy\u003c/h2\u003e \u003cp\u003ePrediction accuracy was evaluated by comparing the prosthesis type selected intraoperatively with the preoperatively planned model. The AI-based planning group demonstrated higher prediction accuracy rates for both the femoral side (79.5%) and tibial side (84.1%). In contrast, the conventional 2D X-ray templating group showed lower accuracy rates of 52.3% for the femoral component and 61.4% for the tibial component. The differences in accuracy between the two groups were statistically significant (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Detailed information are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePrediction accuracy\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraditional group (44)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrediction accuracy of the femoral side\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.023\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAccurate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e35(79.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23(52.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGeneral\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8(18.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e17(38.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePoor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1(2.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4(9.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrediction accuracy of the tibial side\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAccurate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e37(84.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e27(61.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGeneral\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6(13.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14(31.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePoor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1(2.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3(6.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Perioperative Indicators\u003c/h2\u003e \u003cp\u003ePerioperative outcomes are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The mean operative time was 118.45\u0026thinsp;\u0026plusmn;\u0026thinsp;27.21 minutes for the AI-robotic group and 112.86\u0026thinsp;\u0026plusmn;\u0026thinsp;25.28 minutes for the traditional group, with no statistically significant difference between groups (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). However, the AI-robotic group demonstrated significantly shorter osteotomy time (15.24\u0026thinsp;\u0026plusmn;\u0026thinsp;4.71 minutes vs. 18.43\u0026thinsp;\u0026plusmn;\u0026thinsp;4.76 minutes, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.002), reduced intraoperative blood loss (197.41\u0026thinsp;\u0026plusmn;\u0026thinsp;78.41 mL vs. 234.35\u0026thinsp;\u0026plusmn;\u0026thinsp;74.53 mL, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.026), and lower total postoperative blood loss within 72 hours (1022.96\u0026thinsp;\u0026plusmn;\u0026thinsp;226.14 mL vs. 1118.71\u0026thinsp;\u0026plusmn;\u0026thinsp;193.30 mL, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.036) compared to the traditional group. These findings indicate that AI-assisted planning combined with robotic surgery significantly improved intraoperative efficiency and reduced perioperative blood loss (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePeroperative indicators\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraditional group (44)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTime of operation (min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e118.61\u0026thinsp;\u0026plusmn;\u0026thinsp;27.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e112.86\u0026thinsp;\u0026plusmn;\u0026thinsp;25.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.307\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTime of osteotomy (min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e15.24\u0026thinsp;\u0026plusmn;\u0026thinsp;4.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e18.43\u0026thinsp;\u0026plusmn;\u0026thinsp;4.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePeroperative bleeding (ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e197.41\u0026thinsp;\u0026plusmn;\u0026thinsp;78.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e234.35\u0026thinsp;\u0026plusmn;\u0026thinsp;74.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal blood loss (ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e1022.96\u0026thinsp;\u0026plusmn;\u0026thinsp;226.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e1118.71\u0026thinsp;\u0026plusmn;\u0026thinsp;193.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Postoperative Radiographic Outcomes\u003c/h2\u003e \u003cp\u003eDetailed measurements are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. In the AI-robotic group, the postoperative HKA angle was 178.21\u0026thinsp;\u0026plusmn;\u0026thinsp;1.10\u0026deg;, with a malalignment rate of 11.4%. The FFC, FTC, LFC, LTC angle were 84.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.92\u0026deg;, 88.22\u0026thinsp;\u0026plusmn;\u0026thinsp;0.88\u0026deg;, 5.87\u0026thinsp;\u0026plusmn;\u0026thinsp;2.18\u0026deg;, and 84.71\u0026thinsp;\u0026plusmn;\u0026thinsp;2.48\u0026deg;, respectively. Between-group comparisons demonstrated a statistically significant difference in the LFC angle (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while no significant differences were observed in the HKA angle, HKA outlier rate, FFC, FTC, or LTC (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePostoperative imaging data\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGroup\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePostoperative HKA(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{x}\\)\u003c/span\u003e\u003c/span\u003e\u0026plusmn;s)/\u0026deg;\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHKA outlier (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFFC(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{x}\\)\u003c/span\u003e\u003c/span\u003e\u0026plusmn;s)/\u0026deg;\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFTC(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{x}\\)\u003c/span\u003e\u003c/span\u003e\u0026plusmn;s)/\u0026deg;\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eLFC(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{x}\\)\u003c/span\u003e\u003c/span\u003e\u0026plusmn;s)/\u0026deg;\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eLTC(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{x}\\)\u003c/span\u003e\u003c/span\u003e\u0026plusmn;s)/\u0026deg;\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e178.21\u0026thinsp;\u0026plusmn;\u0026thinsp;1.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5(11.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e88.22\u0026thinsp;\u0026plusmn;\u0026thinsp;0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5.87\u0026thinsp;\u0026plusmn;\u0026thinsp;2.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e84.71\u0026thinsp;\u0026plusmn;\u0026thinsp;2.48\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTraditional group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e178.36\u0026thinsp;\u0026plusmn;\u0026thinsp;0.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6(13.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84.00\u0026thinsp;\u0026plusmn;\u0026thinsp;1.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e88.16\u0026thinsp;\u0026plusmn;\u0026thinsp;1.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e6.91\u0026thinsp;\u0026plusmn;\u0026thinsp;2.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e85.03\u0026thinsp;\u0026plusmn;\u0026thinsp;2.19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.488\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.747\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.803\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.530\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Functional Follow-Up Outcomes\u003c/h2\u003e \u003cp\u003ePostoperative functional recovery outcomes are detailed in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. Comparative analysis of follow-up data between groups revealed that at 6 weeks postoperatively, the AI-robotic group exhibited significantly lower Visual Analog Scale (VAS) pain scores (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) and higher Hospital for Special Surgery (HSS) knee scores (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) compared to the traditional group. However, no statistically significant differences were observed in knee range of motion (ROM) between the two groups at 6 weeks, 3 months, or 6 months postoperatively (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). Additionally, both HSS and VAS scores showed no significant intergroup differences at 3 and 6 months postoperatively (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePostoperative functional recovery indicators\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEvaluation indicators\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGroups\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6 weeks\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3 months\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6 months\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.27\u0026thinsp;\u0026plusmn;\u0026thinsp;1.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.18\u0026thinsp;\u0026plusmn;\u0026thinsp;0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.48\u0026thinsp;\u0026plusmn;\u0026thinsp;0.55\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraditional group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.84\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.55\u0026thinsp;\u0026plusmn;\u0026thinsp;1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.029\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.080\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.183\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eROM (\u0026deg;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e106.03\u0026thinsp;\u0026plusmn;\u0026thinsp;2.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e115.47\u0026thinsp;\u0026plusmn;\u0026thinsp;4.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e122.71\u0026thinsp;\u0026plusmn;\u0026thinsp;4.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraditional group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e105.07\u0026thinsp;\u0026plusmn;\u0026thinsp;2.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e114.10\u0026thinsp;\u0026plusmn;\u0026thinsp;4.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e121.41\u0026thinsp;\u0026plusmn;\u0026thinsp;5.22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.125\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.215\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHSS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-robotic group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e61.57\u0026thinsp;\u0026plusmn;\u0026thinsp;4.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e81.91\u0026thinsp;\u0026plusmn;\u0026thinsp;2.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e92.09\u0026thinsp;\u0026plusmn;\u0026thinsp;2.40\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraditional group (44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e59.59\u0026thinsp;\u0026plusmn;\u0026thinsp;3.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e80.89\u0026thinsp;\u0026plusmn;\u0026thinsp;2.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e91.16\u0026thinsp;\u0026plusmn;\u0026thinsp;3.19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.027\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.097\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.125\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003ePreoperative planning is a pivotal component in achieving precision-oriented TKA, as the quality of postoperative functional recovery directly hinges on the accuracy of prosthesis size matching and three-dimensional spatial alignment. Prosthesis size discrepancies can induce abnormal stress distribution, consequently leading to postoperative pain and functional impairment \u003csup\u003e[5]\u003c/sup\u003e. Research indicates that oversized prostheses carry risks of patellar overstuffing and excessive tension in bilateral collateral ligaments, whereas undersized prostheses compromise knee flexion stability and increase the likelihood of component loosening \u003csup\u003e[6]\u003c/sup\u003e. Although conventional X-ray templating remains widely used in clinical practice due to its operational simplicity, low technical threshold, and cost-effectiveness \u003csup\u003e[7]\u003c/sup\u003e, its limitations\u0026mdash;including magnification distortion in two-dimensional radiographs, insufficient anatomical detail, lack of soft tissue-bony landmark visualization, and cumulative manual measurement errors\u0026mdash;may result in significant discrepancies between preoperative plans and intraoperative anatomical realities \u003csup\u003e[8, 9]\u003c/sup\u003e. Arora \u003cem\u003eet al.\u003c/em\u003e \u003csup\u003e[10]\u003c/sup\u003edemonstrated that preoperative planning with X-ray templating for TKA resulted in prosthesis size mismatches in 46.8% of cases, with interobserver and intraobserver variability rates of 46.8% and 43.6%, respectively. These findings highlight the observer-dependent and subjective nature of X-ray templating, which lacks both accuracy and reproducibility. Furthermore, traditional TKA preoperative planning is influenced not only by surgeon subjectivity but also by patient-specific variations \u003csup\u003e[11]\u003c/sup\u003e. Factors such as limb rotation/flexion during X-ray imaging, improper positioning, concomitant knee dysplasia, severe varus/valgus deformities, or joint instability often obscure bony landmarks and introduce measurement errors. In contrast, AI-based preoperative planning overcomes these limitations by generating precise 3D knee models from preoperative CT scans, enhancing visualization of anatomical landmarks and mitigating the drawbacks of conventional methods \u003csup\u003e[12]\u003c/sup\u003e. The AI system intelligently identifies anatomical landmarks, measures critical parameters (e.g., optimal osteotomy thickness), and matches prosthesis sizes from a digital database. Additionally, it simulates surgical outcomes in a 3D virtual environment preoperatively, enabling surgeons to refine plans, improve precision, reduce postoperative complications, extend prosthesis longevity, and enhance patient satisfaction \u003csup\u003e[13]\u003c/sup\u003e. Current studies have demonstrated the significant clinical utility of AI-assisted preoperative planning in total hip arthroplasty (THA) prosthesis sizing \u003csup\u003e[14, 15]\u003c/sup\u003e, but research on its application in TKA remains limited. The AI KNEE preoperative planning system used in this study integrates preoperative CT data with deep learning algorithms to optimize TKA planning. Results showed significantly higher accuracy rates for femoral (79.5% vs. 52.3%) and tibial (84.1% vs. 61.4%) component size prediction in the AI-assisted group compared with conventional 2D templating. Furthermore, the AI system provided 3D visualization and multi-parameter analysis (e.g., femoral valgus angle, tibial posterior slope), offering surgeons multidimensional decision support to minimize intraoperative uncertainties. While some studies suggest that a\u0026thinsp;\u0026plusmn;\u0026thinsp;1 size discrepancy is clinically acceptable for conventional templating\u0026mdash;under which traditional X-ray methods achieve moderate match rates\u0026mdash;this tolerance reflects an inherent compromise of legacy techniques. In contrast, AI-assisted systems enhance prosthesis matching precision, particularly in complex cases (e.g., severe bone defects, joint deformities), addressing the stringent anatomical adaptability requirements of precision and personalized medicine. Using a strict \"zero-size deviation\" criterion (\u0026plusmn;\u0026thinsp;0), our study confirmed that the AI system demonstrated superior accuracy with statistically significant advantages over conventional methods (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). These findings establish AI-assisted preoperative planning as a transformative approach for advancing surgical precision and technical innovation in TKA.\u003c/p\u003e \u003cp\u003ePrecise preoperative planning requires equally precise intraoperative execution to achieve optimal outcomes. The standardized, systematic framework of TKA procedures makes it particularly amenable to robotic-assisted technologies. Studies by Sires et al. \u003csup\u003e[16]\u003c/sup\u003e demonstrated that the Mako robotic system achieves high accuracy in TKA, with postoperative lower limb alignment deviations consistently within \u0026plusmn;\u0026thinsp;3\u0026deg;. Similarly, Zhang et al. \u003csup\u003e[17]\u003c/sup\u003e reported that robotic-assisted TKA enables more accurate component positioning and improved intraoperative joint gap balancing compared with conventional techniques, leading to superior short-term patient outcomes. Additionally, robotic-assisted TKA holds advantages in enhancing surgical proficiency, managing operative pressure, and boosting surgeon confidence, thereby shortening the learning curve. Fu et al. \u003csup\u003e[18]\u003c/sup\u003e observed that patients undergoing robotic-assisted TKA exhibited higher postoperative HSS and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores, improved hip-knee-ankle (HKA) angles, and reduced operative times in follow-up exceeding six months. However, some studies note that robotic-assisted TKA may prolong operative times due to additional steps for robotic frame installation, registration protocols, and the initial learning curve \u003csup\u003e[19\u0026ndash;21]\u003c/sup\u003e. In this study, data showed no significant difference in overall operative time between the robotic-assisted and conventional groups (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.307). However, the osteotomy phase was significantly shorter in the robotic group (15.24\u0026thinsp;\u0026plusmn;\u0026thinsp;4.71 minutes vs. 18.43\u0026thinsp;\u0026plusmn;\u0026thinsp;4.76 minutes, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.002). This efficiency advantage may be attributed to four mechanisms: (1) Precision-Driven Osteotomy Minimizes Revisions: The robotic system enables high-accuracy osteotomy (error\u0026thinsp;\u0026lt;\u0026thinsp;1\u0026deg;) through real-time bone volume monitoring (precision: 0.1 mm) and dynamic force feedback control, eliminating time-consuming intraoperative adjustments required in conventional techniques. (2) Seamless Preoperative-Intraoperative Data Integration: The 3D osteotomy plan generated by AI preoperative planning\u0026mdash;including osteotomy volume, angles, and prosthesis sizing\u0026mdash;was directly imported into the robotic navigation system, reducing redundant intraoperative anatomical landmark measurements. (3) Automated Osteotomy Path Execution: The robotic arm performed rapid bone resections along predefined trajectories, outperforming manual oscillating saws (which rely on surgeon experience). (4) Standardized Decision-Making Workflow: Augmented reality interfaces visually displayed osteotomy safety margins, simplifying surgical decision-making and minimizing delays from hesitation or misjudgment. Notably, while osteotomy efficiency improved, the additional time required for manual anatomical landmark registration and robotic system calibration partially offset these gains, resulting in comparable overall operative times between groups (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). Future advancements in human-robot interface design and registration algorithms are expected to further reduce total procedural duration.\u003c/p\u003e \u003cp\u003eAs a major orthopedic intervention, TKA is associated with substantial blood loss risks and subsequent transfusion-related complications, including immune reactions, infections, and disease transmission \u003csup\u003e[22]\u003c/sup\u003e, and studies report transfusion rates as high as 19% in TKA \u003csup\u003e[23]\u003c/sup\u003e, underscoring the critical need to mitigate bleeding risks in joint surgery. In this study, the robotic-assisted group demonstrated significantly lower intraoperative blood loss (197.41\u0026thinsp;\u0026plusmn;\u0026thinsp;78.41 mL vs. 234.35\u0026thinsp;\u0026plusmn;\u0026thinsp;74.53 mL, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.026) and reduced total postoperative blood loss within 72 hours (1,022.96\u0026thinsp;\u0026plusmn;\u0026thinsp;226.14 mL vs. 1,118.71\u0026thinsp;\u0026plusmn;\u0026thinsp;193.30 mL, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.036) compared with the conventional group, a hemostatic advantage attributed to the elimination of femoral intramedullary canal access\u0026mdash;a mandatory step in traditional TKA for establishing the femoral mechanical axis but linked to medullary cavity trauma and heightened intraoperative bleeding\u0026mdash;and to robotic-assisted TKA using mechanical arm navigation for osteotomy, which obviates medullary cavity penetration and reduces intraoperative hemorrhage. Additionally, the shorter osteotomy duration in the robotic group likely contributed to blood loss reduction by minimizing bone surface exposure time. While Khan \u003cem\u003eet al\u003c/em\u003e. \u003csup\u003e[24]\u003c/sup\u003e proposed that robot-assisted TKA reduces intraoperative blood loss and transfusion rates via minimized soft tissue release and precise osteotomy, Held \u003cem\u003eet al.\u003c/em\u003e \u003csup\u003e[25]\u003c/sup\u003e demonstrated significantly greater blood loss in robot-assisted TKA compared with conventional techniques when assessed via pre-/postoperative hematocrit and hemoglobin dynamics, highlighting the need for prospective studies with extended follow-up to validate the long-term hematoprotective effects of robotic technology and address interstudy discrepancies stemming from methodological variations in blood loss quantification and heterogeneous surgical protocols.\u003c/p\u003e \u003cp\u003eIn the imaging domain, this study revealed that patients undergoing TKA with AI-assisted preoperative planning and robotic navigation exhibited significantly more optimal LFC angles compared with the conventional group (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). The LFC angle, a critical indicator of osteotomy precision, reflects the positional relationship between the femoral prosthesis and the distal femoral anterior cortex. A smaller LFC angle may indicate intraoperative osteotomy notching, risking femoral fractures or prosthesis anterior tilt, whereas a larger angle may cause patellar overstuffing and postoperative pain \u003csup\u003e[26, 27]\u003c/sup\u003e. The AI-Robotic Group's improved LFC angle stemmed from three mechanisms: (1) AI planning using patient-specific CT to identify distal femoral anterior cortex notching points; (2) intraoperative registration verifying preoperative notching references to enhance accuracy; (3) robotic arm-assisted osteotomy with real-time volume monitoring for precise bone resection. Additionally, the AI-Robotic Group achieved postoperative hip-knee-ankle (HKA) angles closer to the ideal 180\u0026deg; mechanical alignment with a lower deviation rate, though statistical significance was not reached (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003eThe follow-up results of this study demonstrated that the AI-robotic group exhibited superior VAS pain scores (2.27\u0026thinsp;\u0026plusmn;\u0026thinsp;1.12 vs. 2.84\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.029) and HSS scores (61.57\u0026thinsp;\u0026plusmn;\u0026thinsp;4.40 vs. 59.59\u0026thinsp;\u0026plusmn;\u0026thinsp;3.80, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.027) compared to the conventional group at 6 weeks postoperatively, with this early functional advantage potentially stemming from multifactorial synergy: Firstly, the AI preoperative planning system precisely analyzed CT data to intelligently match the optimal prosthesis size while optimizing osteotomy surfaces, achieving anatomical alignment between the prosthesis geometry and the resected bone with higher surface coverage to reduce synovial inflammatory reactions caused by micromotion friction; second, the robotic system dynamically monitored flexion-extension gap pressure differences intraoperatively, adjusting osteotomy volume in real-time based on mechanical feedback to avoid excessive soft tissue release due to compensatory osteotomy errors common in traditional surgery; additionally, patient-specific joint line reconstruction based on 3D modeling accurately restored patellofemoral tracking, reducing contact pressure and consequently diminishing pain associated with abnormal mechanical loading; notably, precision osteotomy techniques also decreased bone debris release, potentially suppressing β-glycerophosphate phosphatase and proinflammatory cytokines (IL-6, TNF-α) expression levels to modulate the local inflammatory microenvironment and accelerate postoperative pain relief. However, the convergence of functional and pain scores between groups at 3 and 6 months postoperatively suggests that the core advantages of AI robotic technology may primarily focus on perioperative tissue protection and early biomechanical optimization, while long-term functional recovery depends more on patients' inherent rehabilitation capacity and soft tissue remodeling capabilities. Existing studies on robot-assisted TKA show partial consensus and differences, with Khlopas et al. \u003csup\u003e[28]\u003c/sup\u003e finding significantly better knee function (especially complex mobility) within 3 months postoperatively with Mako robot-assisted TKA than with conventional surgery, Xu et al.'s RCT \u003csup\u003e[29]\u003c/sup\u003e demonstrating superior tibial prosthesis alignment in the robotic group at 3 months postoperatively but noting that its correlation with functional scores or prosthesis survival rates still requires longer follow-up for validation, and long-term studies by Cho et al. \u003csup\u003e[30]\u003c/sup\u003e and Lee et al \u003csup\u003e[31]\u003c/sup\u003e. (average follow-up of 10\u0026ndash;11 years) indicating that the robotic group had advantages in maintaining mechanical alignment and prosthesis positioning accuracy though differences in functional scores were not statistically significant, suggesting that imaging advantages may take longer to translate into improved clinical outcomes. The results of this study show better knee function scores within 6 weeks postoperatively with AI robot-assisted TKA but only confirm short-term benefits, with further extension of follow-up time needed to evaluate the mid- to long-term efficacy of robot-assisted TKA and determine whether the personalized alignment, reduced soft tissue release, and precision osteotomy achieved with AI robot assistance can translate into sustained postoperative outcomes.\u003c/p\u003e \u003cp\u003eThis study was a retrospective analysis. Although inherent selection bias risks existed, strict inclusion/exclusion criteria were prospectively defined preoperatively, and baseline data showed no significant statistical differences between groups in age, sex, BMI, and preoperative functional scores. All procedures were conducted by a single senior surgical team using a consistent prosthesis brand and model, with postoperative care delivered by a dedicated specialized nursing team following standardized protocols. These measures collectively mitigated selection bias risks to the fullest extent achievable.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eAI-assisted preoperative planning with robotic TKA improves prosthesis sizing accuracy, reduces perioperative blood loss and 72h total blood loss, and enhances early functional outcomes compared to conventional methods. These findings support AI-robotic integration as a precision solution for TKA, particularly in complex cases.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eC\u003c/strong\u003e\u003cstrong\u003eonsent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis retrospective study was approved by the Ethics Committee of Beijing Shijitan Hospital, Capital Medical University (Ethics approval No: sjtkyll-lx-2022(87)) and conducted in accordance with the principles of the Declaration of Helsinki. Given the retrospective nature of the research and the use of de-identified patient data, the Ethics Committee waived the requirement for individual informed consent.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo funding was received for this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYingdong Hu: acquired data, analyzed and interpreted data, and wrote the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHongxing Song: analyzed and interpreted data, and wrote the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYuyu Fan: analyzed and interpreted data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eZerui Sun: acquired data.\u003c/p\u003e\n\u003cp\u003eAll authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":" References","content":"\u003col\u003e\n\u003cli\u003eFeng B, Zhu W, Bian YY, Chang X, Cheng KY, Weng XS: China artificial joint annual data report. \u003cem\u003eChin Med J (Engl) \u003c/em\u003e2020, 134(6):752-753.\u003c/li\u003e\n\u003cli\u003evon Eisenhart-Rothe R, Hinterwimmer F, Graichen H, Hirschmann MT: Artificial intelligence and robotics in TKA surgery: promising options for improved outcomes? \u003cem\u003eKnee Surg Sports Traumatol Arthrosc \u003c/em\u003e2022, 30(8):2535-2537.\u003c/li\u003e\n\u003cli\u003eNadler SB, Hidalgo JH, Bloch T: Prediction of blood volume in normal human adults. \u003cem\u003eSurgery \u003c/em\u003e1962, 51(2):224-232.\u003c/li\u003e\n\u003cli\u003eGross JB: Estimating allowable blood loss: corrected for dilution. \u003cem\u003eAnesthesiology \u003c/em\u003e1983, 58(3):277-280.\u003c/li\u003e\n\u003cli\u003eLan Q, Li S, Zhang J, Guo H, Yan L, Tang F: Reliable prediction of implant size and axial alignment in AI-based 3D preoperative planning for total knee arthroplasty. \u003cem\u003eSci Rep \u003c/em\u003e2024, 14(1):16971.\u003c/li\u003e\n\u003cli\u003eTang A, Yeroushalmi D, Zak S, Lygrisse K, Schwarzkopf R, Meftah M: The effect of implant size difference on patient outcomes and failure after bilateral simultaneous total knee arthroplasty. \u003cem\u003eJ Orthop \u003c/em\u003e2020, 22:282-287.\u003c/li\u003e\n\u003cli\u003eHsu AR, Gross CE, Bhatia S, Levine BR: Template-directed instrumentation in total knee arthroplasty: cost savings analysis. \u003cem\u003eOrthopedics \u003c/em\u003e2012, 35(11):e1596-1600.\u003c/li\u003e\n\u003cli\u003eConn KS, Clarke MT, Hallett JP: A simple guide to determine the magnification of radiographs and to improve the accuracy of preoperative templating. \u003cem\u003eJ Bone Joint Surg Br \u003c/em\u003e2002, 84(2):269-272.\u003c/li\u003e\n\u003cli\u003eBoese CK, Bredow J, Dargel J, Eysel P, Geiges H, Lechler P: Calibration Marker Position in Digital Templating of Total Hip Arthroplasty. \u003cem\u003eJ Arthroplasty \u003c/em\u003e2016, 31(4):883-887.\u003c/li\u003e\n\u003cli\u003eArora J, Sharma S, Blyth M: The role of pre-operative templating in primary total knee replacement. \u003cem\u003eKnee Surg Sports Traumatol Arthrosc \u003c/em\u003e2005, 13(3):187-189.\u003c/li\u003e\n\u003cli\u003eGoyal N, Stulberg SD: Evaluating the Precision of Preoperative Planning in Patient Specific Instrumentation: Can a Single MRI Yield Different Preoperative Plans? \u003cem\u003eJ Arthroplasty \u003c/em\u003e2015, 30(7):1250-1253.\u003c/li\u003e\n\u003cli\u003eSiddiqi A, Hardaker WM, Eachempati KK, Sheth NP: Advances in Computer-Aided Technology for Total Knee Arthroplasty. \u003cem\u003eOrthopedics \u003c/em\u003e2017, 40(6):338-352.\u003c/li\u003e\n\u003cli\u003eJones CW, Jerabek SA: Current Role of Computer Navigation in Total Knee Arthroplasty. \u003cem\u003eJ Arthroplasty \u003c/em\u003e2018, 33(7):1989-1993.\u003c/li\u003e\n\u003cli\u003eChen X, Liu X, Wang Y, Ma R, Zhu S, Li S, Li S, Dong X, Li H, Wang G\u003cem\u003e et al\u003c/em\u003e: Development and Validation of an Artificial Intelligence Preoperative Planning System for Total Hip Arthroplasty. \u003cem\u003eFront Med (Lausanne) \u003c/em\u003e2022, 9:841202.\u003c/li\u003e\n\u003cli\u003eHuo J, Huang G, Han D, Wang X, Bu Y, Chen Y, Cai D, Zhao C: Value of 3D preoperative planning for primary total hip arthroplasty based on artificial intelligence technology. \u003cem\u003eJ Orthop Surg Res \u003c/em\u003e2021, 16(1):156.\u003c/li\u003e\n\u003cli\u003eSires JD, Craik JD, Wilson CJ: Accuracy of Bone Resection in MAKO Total Knee Robotic-Assisted Surgery. \u003cem\u003eJ Knee Surg \u003c/em\u003e2021, 34(7):745-748.\u003c/li\u003e\n\u003cli\u003eZhang J, Ndou WS, Ng N, Gaston P, Simpson PM, Macpherson GJ, Patton JT, Clement ND: Robotic-arm assisted total knee arthroplasty is associated with improved accuracy and patient reported outcomes: a systematic review and meta-analysis. \u003cem\u003eKnee Surg Sports Traumatol Arthrosc \u003c/em\u003e2022, 30(8):2677-2695.\u003c/li\u003e\n\u003cli\u003eFu X, She Y, Jin G, Liu C, Liu Z, Li W, Jin R: Comparison of robotic-assisted total knee arthroplasty: an updated systematic review and meta-analysis. \u003cem\u003eJ Robot Surg \u003c/em\u003e2024, 18(1):292.\u003c/li\u003e\n\u003cli\u003eVermue H, Luyckx T, Winnock de Grave P, Ryckaert A, Cools AS, Himpe N, Victor J: Robot-assisted total knee arthroplasty is associated with a learning curve for surgical time but not for component alignment, limb alignment and gap balancing. \u003cem\u003eKnee Surg Sports Traumatol Arthrosc \u003c/em\u003e2022, 30(2):593-602.\u003c/li\u003e\n\u003cli\u003eNaziri Q, Cusson BC, Chaudhri M, Shah NV, Sastry A: Making the transition from traditional to robotic-arm assisted TKA: What to expect? A single-surgeon comparative-analysis of the first-40 consecutive cases. \u003cem\u003eJ Orthop \u003c/em\u003e2019, 16(4):364-368.\u003c/li\u003e\n\u003cli\u003eSodhi N, Khlopas A, Piuzzi NS, Sultan AA, Marchand RC, Malkani AL, Mont MA: The Learning Curve Associated with Robotic Total Knee Arthroplasty. \u003cem\u003eJ Knee Surg \u003c/em\u003e2018, 31(1):17-21.\u003c/li\u003e\n\u003cli\u003eCharoencholvanich K, Siriwattanasakul P: Tranexamic acid reduces blood loss and blood transfusion after TKA: a prospective randomized controlled trial. \u003cem\u003eClin Orthop Relat Res \u003c/em\u003e2011, 469(10):2874-2880.\u003c/li\u003e\n\u003cli\u003eMa Q, Sun CJ, Wu S, Cai X: Comparison of Blood Loss between Open-Box Prosthesis and Closed-Box Prosthesis after Primary Total Knee Arthroplasty. \u003cem\u003eOrthop Surg \u003c/em\u003e2021, 13(3):768-777.\u003c/li\u003e\n\u003cli\u003eKhan H, Dhillon K, Mahapatra P, Popat R, Zakieh O, Kim WJ, Nathwani D: Blood loss and transfusion risk in robotic-assisted knee arthroplasty: A retrospective analysis. \u003cem\u003eInt J Med Robot \u003c/em\u003e2021, 17(6):e2308.\u003c/li\u003e\n\u003cli\u003eHeld MB, Gazgalis A, Neuwirth AL, Shah RP, Cooper HJ, Geller JA: Imageless robotic-assisted total knee arthroplasty leads to similar 24-month WOMAC scores as compared to conventional total knee arthroplasty: a retrospective cohort study. \u003cem\u003eKnee Surg Sports Traumatol Arthrosc \u003c/em\u003e2022, 30(8):2631-2638.\u003c/li\u003e\n\u003cli\u003eJin Z, Wang Z, Xu K, Chu J, Xiang S, Tang Y, Wang R, Hua H, Zhang Z, Tong P\u003cem\u003e et al\u003c/em\u003e: Effect of anterior femoral cortical notch grade on postoperative function and complications during TKA surgery: A multicenter, retrospective study. \u003cem\u003eOpen Med (Wars) \u003c/em\u003e2024, 19(1):20240932.\u003c/li\u003e\n\u003cli\u003eOkamoto Y, Nakajima M, Jotoku T, Otsuki S, Neo M: Capsular release around the intercondylar notch increases the extension gap in posterior-stabilized rotating-platform total knee arthroplasty. \u003cem\u003eKnee \u003c/em\u003e2016, 23(4):730-735.\u003c/li\u003e\n\u003cli\u003eKhlopas A, Sodhi N, Hozack WJ, Chen AF, Mahoney OM, Kinsey T, Orozco F, Mont MA: Patient-Reported Functional and Satisfaction Outcomes after Robotic-Arm-Assisted Total Knee Arthroplasty: Early Results of a Prospective Multicenter Investigation. \u003cem\u003eJ Knee Surg \u003c/em\u003e2020, 33(7):685-690.\u003c/li\u003e\n\u003cli\u003eXu J, Li L, Fu J, Xu C, Ni M, Chai W, Hao L, Zhang G, Chen J: Early Clinical and Radiographic Outcomes of Robot-Assisted Versus Conventional Manual Total Knee Arthroplasty: A Randomized Controlled Study. \u003cem\u003eOrthop Surg \u003c/em\u003e2022, 14(9):1972-1980.\u003c/li\u003e\n\u003cli\u003eCho KJ, Seon JK, Jang WY, Park CG, Song EK: Robotic versus conventional primary total knee arthroplasty: clinical and radiological long-term results with a minimum follow-up of ten years. \u003cem\u003eInt Orthop \u003c/em\u003e2019, 43(6):1345-1354.\u003c/li\u003e\n\u003cli\u003eLee YM, Kim GW, Lee CY, Song EK, Seon JK: No Difference in Clinical Outcomes and Survivorship for Robotic, Navigational, and Conventional Primary Total Knee Arthroplasty with a Minimum Follow-up of 10 Years. \u003cem\u003eClin Orthop Surg \u003c/em\u003e2023, 15(1):82-91.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-musculoskeletal-disorders","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmsd","sideBox":"Learn more about [BMC Musculoskeletal Disorders](http://bmcmusculoskeletdisord.biomedcentral.com/)","snPcode":"","submissionUrl":"https://author-welcome.nature.com/12891","title":"BMC Musculoskeletal Disorders","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Total knee replacement surgery, Artificial Intelligence, Robotic-assisted surgery, Preoperative planning","lastPublishedDoi":"10.21203/rs.3.rs-6812822/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6812822/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjective:\u003c/strong\u003e To compare perioperative outcomes and early functional recovery between AI-robotic and conventional total knee arthroplasty (TKA).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eWe retrospectively analyzed data from 88 patients who underwent primary unilateral TKA for knee osteoarthritis between April 2024 and December 2024. The AI-robotic group (n=44) received AI-assisted preoperative planning and robot-assisted TKA, while the traditional group (n=44) underwent conventional 2D templating and manual TKA. Key metrics included preoperative prosthesis prediction accuracy, intraoperative and postoperative blood loss, osteotomy time, postoperative radiographic alignment, and functional scores.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eThe AI-robotic group showed significantly higher prosthesis prediction accuracy (femoral: 79.5% vs. 52.3%, \u003cem\u003eP\u003c/em\u003e=0.023; tibial: 84.1% vs. 61.4%, \u003cem\u003eP\u003c/em\u003e=0.042), shorter osteotomy time (15.24±4.71 vs. 18.43±4.76 minutes, \u003cem\u003eP\u003c/em\u003e=0.002), reduced intraoperative blood loss (197.41±78.41 vs. 234.35±74.53 mL, \u003cem\u003eP\u003c/em\u003e=0.026), and lower 72-hour total blood loss (1022.96±226.14 vs. 1118.71±193.30 mL, \u003cem\u003eP\u003c/em\u003e=0.036). Postoperative lateral femoral component (LFC) angles were superior in the AI-robotic group (5.87±2.18° vs. 6.91±2.10°, \u003cem\u003eP\u003c/em\u003e=0.025). At 6 weeks, the AI-robotic group had better VAS pain scores (2.27±1.12 vs. 2.84±1.22, \u003cem\u003eP\u003c/em\u003e=0.029) and HSS scores (61.57±4.40 vs. 59.59±3.80, \u003cem\u003eP\u003c/em\u003e=0.027).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion: \u003c/strong\u003eAI-assisted preoperative planning with robotic TKA improves prosthesis sizing accuracy, reduces perioperative blood loss and 72h total blood loss, and enhances early functional outcomes compared to conventional methods. These findings support AI-robotic integration as a precision solution for TKA, particularly in complex cases.\u003c/p\u003e","manuscriptTitle":"AI-assisted Preoperative Planning Combined with Robotic-assisted Total Knee Arthroplasty vs. Conventional Surgery: A Retrospective Controlled Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-18 19:32:36","doi":"10.21203/rs.3.rs-6812822/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-22T06:59:08+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-19T10:22:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"146978343953469082615849619911328734637","date":"2025-07-13T19:34:32+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-13T15:48:51+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"105028036142612203072709286373417076631","date":"2025-07-13T15:15:30+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-07T21:22:20+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"228241874001638610769164263646569285062","date":"2025-06-23T15:52:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"254924006677133755558640189638308399643","date":"2025-06-22T13:37:19+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"325130160866585799008224824291205585159","date":"2025-06-15T22:25:38+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-13T08:21:19+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-06-05T18:09:30+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-05T04:30:26+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-05T04:27:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Musculoskeletal Disorders","date":"2025-06-03T15:15:53+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-musculoskeletal-disorders","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmsd","sideBox":"Learn more about [BMC Musculoskeletal Disorders](http://bmcmusculoskeletdisord.biomedcentral.com/)","snPcode":"","submissionUrl":"https://author-welcome.nature.com/12891","title":"BMC Musculoskeletal Disorders","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"36e5e44f-74a6-4644-9367-4cad6c058286","owner":[],"postedDate":"June 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-01-19T16:50:17+00:00","versionOfRecord":{"articleIdentity":"rs-6812822","link":"https://doi.org/10.1186/s12891-025-09455-5","journal":{"identity":"bmc-musculoskeletal-disorders","isVorOnly":false,"title":"BMC Musculoskeletal Disorders"},"publishedOn":"2026-01-16 16:30:59","publishedOnDateReadable":"January 16th, 2026"},"versionCreatedAt":"2025-06-18 19:32:36","video":"","vorDoi":"10.1186/s12891-025-09455-5","vorDoiUrl":"https://doi.org/10.1186/s12891-025-09455-5","workflowStages":[]},"version":"v1","identity":"rs-6812822","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6812822","identity":"rs-6812822","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}