Semiautomatic 3D Ultrasound Kidney Volume Segmentation in Pediatric Hydronephrosis: Interrater Agreement and Correlation to Conventional Hydronephrosis Grading

Semiautomatic 3D Ultrasound Kidney Volume Segmentation in Pediatric Hydronephrosis: Interrater Agreement and Correlation to Conventional Hydronephrosis Grading | 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 Semiautomatic 3D Ultrasound Kidney Volume Segmentation in Pediatric Hydronephrosis: Interrater Agreement and Correlation to Conventional Hydronephrosis Grading Michael Esser, Ilias Tsiflikas, James R. Jago, Laurence Rouet, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4868701/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 06 May, 2025 Read the published version in Pediatric Radiology → Version 1 posted 15 You are reading this latest preprint version Abstract Background: Two-dimensional assessment of the kidney volume underestimates real kidney volume with a high interobserver variability. Limited data exist on innovative 3D ultrasound (3DUS) technique for the evaluation of pediatric hydronephrosis. Objective: To evaluate the interrater agreement of kidney volume segmentation by 3DUS with a matrix array transducer in children with hydronephrosis and comparison of 3D metrics to conventional hydronephrosis grading. Materials and methods: 48 kidney volumes were prospectively acquired in 45 patients with hydronephrosis by freehand 3DUS (6-1MHz volumetric sector array, electronic rotation; median age, 4years; 1month to 16years). Semi-automated kidney segmentation was performed by two independent readers providing volumes for total kidney (renal capsule), dilated collective system, renal parenchyma (renal capsule - collective system) and hydronephrosis index (renal parenchyma / renal capsule). Interrater agreement was evaluated with Bland–Altman plots, intraclass correlation coefficient (ICC) and Dice similarity coefficients. The maximum 2D diameter of renal pelvis was measured and hydronephroses were morphologically classified grade 1-4. Results: Interrater agreement for renal capsule, collective system, hydronephrosis index and renal parenchyma was good to excellent with ICC of 0.94, 0.87, 0.83 and 0.92 respectively (p<0.001 each). Median Dice was 0.90 (capsule), 0.77 (collective system) and 0.88 (parenchyma). There was a positive correlation between hydronephrosis grading and ultrasonic hydronephrosis index and between renal pelvis diameter and collective system volume (p<0.001 both). Conclusion: Semiautomatic 3DUS volumetric analysis has a high degree of interrater agreement providing parenchyma volume in hydronephrotic kidneys. Volumes of the collective system and hydronephrosis index correlate with the extent of hydronephrosis. Trial registry: trial registration number, DRKS00022772; date of registration, 07/31/2020 Hydronephrosis Pediatrics Kidney Diseases Imaging Three-Dimensional Ultrasonography Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction In pediatric hydronephrosis the evaluation of renal parenchymal volume plays an important role for the initial estimation of prognosis and for disease monitoring. It deals as a surrogate of renal function and parenchymal damage suggesting the need for invasive diagnostic and therapeutic intervention [ 1 ]. The non-invasive and radiation-free sonographic assessment of renal volume is considered an indicator for the real renal size and nephron mass (renal corpuscles and canaliculi) [ 2 ]. Conventional ultrasound-based assessment of the kidney volume is already well established since decades; however, it is mainly based on two-dimensional (2D) measurements using the greatly simplified ellipsoid formula, which is known to underestimate real kidney volume with a high interobserver variability [ 3 – 5 ]. In fact, using the ellipsoid formula can be understood as a pragmatic approach of estimating renal volume which is easy to calculate in regular non-hydronephrotic kidneys. However, bearing in mind the highly complex anatomy of the renal pelvicalyceal system, further difficulty for accurate renal volume assessment arises with hydronephrosis and possible associated malformation or status after renal surgery. In this context, the severity of pediatric hydronephrosis is a crucial point when considering the dynamic process of a growing kidney and changing degrees of renal pelvic dilation in different clinical situations [ 1 , 3 ]. Anteroposterior diameter with simple thresholds and calyceal dilation by themselves are insufficient parameters as they are affected by many factors like hydration status and bladder filling, and do not demonstrate the real degree of hydronephrosis in different pelvic configurations [ 6 ]. The different radiology grading systems are based on subjective criteria and are thus not recommended for therapeutic decision making. As 2D ultrasound (2DUS) measurements may be inaccurate in anatomically complex kidneys, no gold standard in imaging exists for determining neither the severity of hydronephrosis nor the volume of renal parenchyma [ 1 , 7 ]. Today, three-dimensional ultrasound (3DUS) has become an established imaging tool in some specialties. Since the early 2000s 3DUS of the kidney is increasingly used in the pediatric population with a reported improvement of volume assessment accuracy at a low interobserver variability [ 3 , 7 – 9 ]. Yet, only limited data exist on 3DUS for the evaluation of pediatric hydronephrosis where an added value might be expected compared to regular kidney morphology. Prior studies on renal 3DUS concentrated on the adult population [ 10 – 12 ], they included non-hydronephrotic kidneys [ 10 , 11 , 13 ] and/or were restricted to preselected clinical subgroups [ 8 ]. In the course of the technical progress in ultrasound imaging systems, matrix array transducers have been developed for a nearly instant free-hand acquisition of volumetric data. They can be handled like conventional transducers as they do not need an electric motor or an electromagnetic position sensor. They have shown to considerably reduce volumetric errors caused by motion artifacts (which may occur in children who are not able to hold their breath or only for a short time). Initially, matrix array transducers for renal volume calculation have been evaluated in adults including a cadaver study and comparisons of normal kidney volume to CT and renal function [ 11 , 12 ]. In children a matrix array transducer was used to compare renal volumes in polycystic kidneys [ 8 ]. The only report on a matrix array transducer in hydronephrotic kidneys included eight kidneys for the feasibility study of a new segmentation model [ 14 ]. To the best of our knowledge, no further clinical data exist on the use of a matrix array transducer in hydronephrotic kidneys in children. The purpose of our study was to evaluate the interrater agreement of kidney volume segmentation by three-dimensional ultrasound (3DUS) with a matrix array transducer in children with hydronephrosis and compare the resulting metrics to conventional hydronephrosis grading. Materials and methods This prospective monocenter study was approved by the local ethics committee. Signed informed consent was obtained from all legal guardians and children above the age of 11 participating in the study. All study procedures were conducted in accordance with the Guidelines for Good Clinical Practice and ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. The authors had full control of the data. Study population Between March and September 2021 all patients under 17 years who were referred to our sonography center as part of the in-house pediatric urology consultation with a clinical indication for ultrasound of the kidneys and present hydronephrosis were included in the study. Patients were excluded from the study if patients or legal guardians declined afterwards to participate in the study. 3D Ultrasound Acquisition and Evaluation Initially, routine 2DUS was performed by a radiologist with 7 years of experience in ultrasound using an EPIQ 5G ultrasound device (Philips Medical Systems, Bothell, WA, USA). The maximum anteroposterior diameter of renal pelvis was measured on transverse 2D B-mode images and hydronephroses were classified grade 1–4 following the Onen-2007 hydronephrosis grading system [ 1 ]. Afterwards, freehand 3DUS with a linear matrix array transducer (X6-1) was performed by the same radiologist immediately after 2D sonography (Fig. 1 ). For this purpose, all patients were placed in prone position. First, image parameters were adjusted with the help of protocols which were optimized for children by the manufacturer and stored on the device. With activation of the 3D X-plane mode, a live side-by-side-image included simultaneous sagittal and axial views of the kidney to ensure that the entire kidney will be included in the 3D volume. The acquisition scan of one single data set for each kidney took two to three seconds. The multiplanar reformatted (MPR) display showed the structure in three orthogonal planes as well as a volume-rendered image to check the volume for complete coverage of the kidney. For each kidney three subsequent volumetric data sets were acquired. The 3D data sets were automatically stored on an integrated hard disc. Time for the acquisition of the three data sets was between one and three minutes. Semi-automated 3D kidney segmentation was performed on a separate workstation using research prototype software (Philips Health Technology Innovation, Paris, France). The prototype segmentation is based on implicit shape deformation [ 15 ]. The principle is to deform a 3D kidney template shape according to internal and external forces. The forces are governed by fixed parameters tuned for 3D Ultrasound of kidney volumes. After initial deformation, the reader can apply interactive corrections to adjust the segmentation to the exact kidney capsule borders. In a second step, a region growing algorithm is used to interactively segment the collective system. The collective system segmentation may be composed of multiple sub-volumes. Upon completion, segmentations and corresponding volumes quantifications are saved. Two readers segmented the kidneys freely choosing one of the three saved data sets for each kidney. The readers were instructed to choose a data set with respect to the best coverage of organ surface and with minimal motion artifacts. The readers were blinded to all identifying and clinical data. For renal capsule segmentation a sphere was placed in an axial view to define the center of the kidney. The sphere was used to initialize the size and pose of the kidney 3DUS template shape, which was then deformed. Afterwards, the ellipsoid capsule was adjusted by providing additional marking points in the upper and lower poles of the kidney using a sagittal view. In a separate edition mode, the intra- and extrarenal portions of the (dilated) collective system including calices were segmented. The resulting segmentation volumes were checked in all three orthogonal planes. Quantitative results were shown in a separate table. The following values were calculated: collective system, renal parenchyma, total kidney volume (addition of collective system and renal parenchyma), hydronephrosis index (HI, kidney parenchyma divided by total renal capsule). The extrarenal proportion of the pelvis was not included for the calculation of HI. Statistics For statistical analysis, IBM SPSS Statistics (version 22 for Windows, Ehningen, Germany) was used. The Kolmogorow-Smirnov test was used to study the distribution of quantification data. Continuous variable data are presented as means ± standard deviations. Data that did not follow a normal distribution are presented as median with interquartile ranges (IQR). Bland–Altman plots were used to visually compare measurements of the two readers for the kidney parenchyma. The mean difference and the upper and lower limits of agreement (LoA) were calculated. The limits of agreement were defined as the mean difference ± 1.96 standard deviations of the differences. To assess the interrater agreement, intraclass correlation coefficient (ICC) estimates and their 95% confidence levels (CL) were calculated using an absolute agreement, two-way mixed-effects model [ 16 ]. The ICC was defined as slight (0–0.20), fair (0.21–0.40), moderate (0.41–0.60), substantial (0.61–0.80), and excellent (0.81–1.00). Spearman's rank correlation test was used to analyze the relationship between hydronephrosis grading and the collective system volumes. A p-value less than 0.05 was considered to indicate statistical significance. To evaluate the voxel-wise spatial overlap of the segmentations, Dice similarity coefficients (DSC) were computed between the 22 segmentations of the two readers in which both readers selected the same data set [ 17 ]. The values of DSC range from 0, representing no spatial overlap between two sets of segmentations, to 1, indicating a complete overlap. Results Patient characteristics 48 kidney volumes were calculated in 45 children (median age 4.5 years; range, 1 month to 16 years; interquartile range, IQR, 7 years; 7 patients under the age of 12 months; 23 patients under the age of 5 years; 35 males, 10 females). The left side (n = 29, 60%) was a little more frequently affected than the right side (n = 19, 40%). The most frequent hydronephrosis grading was grade 2 (n = 29), followed by grade 3 (n = 15). Only 4 patients were included with grade 1 hydronephrosis. The underlying kidney diseases of all patients are summarized in Table 1. In 18/48 kidneys (38%) a prior history of urological surgery was recorded. Interrater Agreement The two readers chose to use the same data set for segmentation in 46% (22/48) of the cases. For an assessment of the range of all kidney parenchyma values the Bland-Altmann-Plot is shown in Fig. 2 . The mean difference was 0.68 ml. The upper limit of agreement was 25.3 ml and the lower limit of agreement was − 24.0 ml. One outlier was identified below the lower limit of agreement: In this 11-year old patient with grade 3 hydronephrosis due to renal infundibular stenosis, parenchyma value difference between the two readers was 65 ml. The collective system was segmented with similar volumes of 39.6 and 40.4. ml, but a relevant discrepancy occurred in renal capsule segmentation (111 vs. 177 ml). In this case, different data sets were used for kidney segmentation. The subsequent manual comparison showed that the upper kidney pole was not fully captured in the reader 1 data set. Interrater agreement between the two readers for total kidney volume (renal capsule; ICC 0.94; 95% CL [0.90; 0.97]; p < 0.001), collective system (ICC 0.87; 95% confidence level (CL) [0.78; 0.93]; p < 0.001), HI (ICC 0.83; 95% CL [0.71; 0.90]; p < 0.001) and renal parenchyma (ICC 0.92; 95% CL [0.86; 0.96]; p < 0.001) was good to excellent. Renal parenchyma values had a median of 62.6 and 62.0 ml for the two readers, ranging from 16.1 ml to 147.7 ml (reader 1; IQR, 37.9 ml) and from 16.5 ml to 139.5 ml (reader 2; IQR 41.3 ml). The HI had a median of 0.91 for both readers, ranging from 0.63 to 0.99 (reader 1; IQR, 0.10) and from 0.52 to 0.98 (reader 2; IQR 0.12). The 22 segmentations with the same data set achieved a median DSC of 0.90 ranging from 0.76 to 0.95 (IQR, 0.06) for the total kidney volume (renal capsule), a median DSC of 0.77 ranging from 0.54 to 0.89 (IQR, 0.76) for the collective system and a median DSC of 0.88 ranging from 0.73 to 0.94 (IQR, 0.10) for the renal parenchyma. The corresponding boxplots for the three DSCs between reader 1 and reader 2 by segmentation pairs are presented in Fig. 3 . In one case a remarkably low DSC for the collective system occurred (DSC, 0.54). In this 8-year old patient with duplex kidney, similar absolute values were obtained for the volumes of the slightly dilated renal pelvis (grade 1; 1.4 vs. 1.5 ml), although the spatial overlap was obviously not optimal. This also resulted in a similar renal volume with a difference between the readers of 4.7 ml (77.2 vs. 72.6 ml). Hydronephrosis Grading Hydronephrosis indices (HI) of both readers were 0.91 (median; IQR, 0.1 for both readers; range for reader 1, 0.63 to 0.99; range for reader 2, 0.52 to 0.98). There was a positive correlation between the hydronephrosis grading and the HI (p < 0.001; Fig. 4 ) and between hydronephrosis grading and the volume of the collective system (p < 0.001). Discussion This study provides initial data concerning the feasibility of a free-hand three-dimensional kidney volume assessment for the evaluation of hydronephrosis in children. The workflow presented here allowed for reproducible and accurate results of volumetric data providing useful information according to established methods of hydronephrosis grading. Measurements can be obtained rapidly and easily, without the use of radiation, contrast agent or anesthesia. The method is thus feasible in neonates, infants and children. The prototype software for semiautomatic segmentation as one component of the presented workflow (Fig. 1 ) allows a reliable subtraction of the dilated collecting system and enables reproducible calculation of renal parenchyma size. A recent review about modern approaches of kidney imaging in children mentioned 3DUS as the “least invasive and most cost-effective way” for repetitive imaging with monitoring of renal size in children and underlined the importance of upcoming investigation on 3DUS as an “important goal for future research” [ 2 ]. As part of this development the first clinical studies reported the use of 3DUS as a promising tool with superior accuracy for the depiction of renal anatomy and reliable parenchyma volume determination in children with renal malformations and for postoperative assessment compared to planimetric analysis [ 7 , 9 ]. In our study, a relevant portion of the cohort had congenital renal abnormalities (Table 1) and/or complex anatomy of the urinary tract, like urogenital sinus or duplex kidney. Renal surgery was another influencing factor that altered renal anatomy in some patients (n = 18, 38%). Therefore, our results meet the former expectations and confirm the favored application of matrix array 3DUS as a robust method in anatomically complex situations. A further study recommended to prefer 3DUS over MRI for accurate kidney volume assessment in the follow-up of children with autosomal dominant polycystic kidney disease (ADPKD) [ 8 ]. In contrast to our study, no hydronephrotic kidneys were evaluated and only older patients (> 8 years) were included. In our study, a relevant proportion of the patients was markedly younger (51% under the age of 5 years) including infants (under the age of 12 months; n = 7) bearing the risk of movement during the 3D volume acquisition. Thus, the fast acquisition by means of a matrix array transducer can assure a reliable volume assessment in a representative patient cohort concerning age distribution. In the comparative study of Fritz et al. infants under the age of 6 months were included, however, additional six infants were excluded from the analysis because of increased motion resulting in unusable image data [ 9 ]. Riccabona et al. also reported that 3DUS was not possible in some uncooperative infants that were excluded from the study [ 3 ]. In our prospective study, no patients were excluded afterwards and at least one of the three acquired data sets was evaluable by the software in all patients. Riccabona et al. assessed hydronephrotic kidneys with an electromagnetic positioning device and a mechanically driven transducer, but did not use an electronic matrix array transducer. [ 3 ]. In the similar study of Fritz et al. only eight patients with hydronephrosis were analyzed [ 9 ]. Both prior studies [ 3 , 9 ] already used semiautomated volume calculation with threshold-based segmentation. In our study a comparable semiautomated segmentation model was used to minimize time requirements. There are some methodological aspects to consider with regard to the semiautomatic segmentation used: Our evaluation has shown that the reliable and exact detection of the kidney boundaries is particularly important for accurate parenchyma assessment. Since the volume of the renal capsule is much larger than that of the renal pelvis, inaccurate measurements of the renal pelvis can also be compensated for. This explains why one case with a large volume deviation of the renal capsule between the readers was an outlier in the comparison of the parenchyma values (Fig. 2 ). The selection of the data set to be segmented could also be important here: In the case mentioned above, two different data sets were segmented. We consider it useful that in routine use with moving children (without the ability to stop breathing), selection from several 3D data can allow greater certainty for accurate subsequent segmentation. However, it only seems to be effective if the renal borders, in particular the upper and lower poles, have been completely captured in all data sets. The time required to save multiple data sets is minimal with the matrix transducer. The initial studies that evaluated matrix array transducers for renal volume calculation in adults compared normal kidney volume to CT [ 12 ] and investigated correlations between renal volume and renal function parameters, but did not evaluate the reproducibility of the technique [ 11 ]. Both studies valued 3DUS with a matrix array transducer as a reliable tool for determining renal volumes with reduced measurement errors and recommended to study the reproducibility in different clinical cohorts [ 12 ]. The only report on a matrix array transducer in hydronephrotic kidneys included eight kidneys for the feasibility study of a new segmentation model [ 14 ]. In this study the renal volumes were analyzed with respect to segmentation errors and relative volume differences, but also no interrater variability was evaluated. In a prior study the hydronephrosis index (HI) has been proposed as a dimensionless parameter of quantifying hydronephrosis in children by 2DUS [ 18 ]. The advantage of standardization in using dimensionless metrics in medical imaging has been mentioned previously [ 19 ]. Rud et al. reported an excellent interobserver agreement for HI and a correlation between HI and the sonographic degree of renal pelvis dilation in adults with stone-related renal colic [ 20 ]. In a prior study, a close correlation between HI and the sonographically evaluated grade of hydronephrosis was shown in patients with pelvi-ureteric junction obstruction [ 21 ]. These study conclusions are in line with our own results with a good interrater agreement (ICC 0.83) and a narrow correlation between HI and the hydronephrosis grading (p < 0.001; Fig. 4 ). Han et al. found that preoperative HI in patients for pyeloplasty can work as a possible prognostic marker for adverse renal function outcome [ 22 ]. The clinical value of 3D ultrasound parameters has not been evaluated in our study. For further studies on HI as an indicator of pediatric renal function in everyday clinical use it has to be kept in mind that 2D sonographic calculation of the index may be time-consuming [ 1 ]. Our results show that semiautomatic calculation of HI in 3D data sets provides a simplified way of objective measurement which is recommendable for further investigations. Our study has the following limitations: In contrast to prior studies we did not perform any comparisons to other imaging modalities (CT or MRI) or to 2DUS. However, the general feasibility and advantages of 3D sonographic evaluations over conventional ultrasound measurements have already been published. A technical limitation is that for all children a transducer with a frequency range of 6 − 1 MHz was used, which resulted in low image resolution in infants and neonates. However, complete coverage of the renal capsule without interpolation of the upper and/or lower pole contour was thus easily possible and image contrast was estimated as suitable for the differentiation of the collective system in all patients. No grade 4 hydronephroses were included, which reduces the generalizability of the results. Conclusion Novel semiautomatic 3DUS volumetric analysis has a high degree of interrater agreement providing parenchyma volume in hydronephrotic kidneys. Volumes of the collective system and hydronephrosis index correlate with the extent of hydronephrosis. Declarations Competing Interests Mr. James R. Jago, PhD is employee of Philips Healthcare. Mrs. Laurence Rouet, PhD is employee of Philips Health Technology Innovation Paris. The study was supported by a research grant in the framework of a collaboration contract with Philips Ultrasound, Inc. Author Contribution M.E.: Conceptualization, Formal analysis and investigation, Original draft preparation, Funding acquisition, SupervisionI.T.: Methodology, Formal analysis and investigation, SupervisionJ.R.J.: Conceptualization, Funding acquisition, ResourcesL.R.: Methodology, Funding acquisition, Resources, Original draft preparationA.S.: Formal analysis and investigation, Original draft preparationJ.F.S.: Conceptualization, Methodology, Funding acquisition, Resources, SupervisionAll authors reviewed the manuscript. Data Availability The data that support the findings of this study are available from Philips Healthcare. Restrictions apply to the availability of these data, which were used under license for this study. References Onen A. Grading of Hydronephrosis: An Ongoing Challenge. Front Pediatr. 2020;8:458. DeFreitas MJ, Katsoufis CP, Infante JC et al. The old becomes new: advances in imaging techniques to assess nephron mass in children. Pediatr Nephrol. 2021;36(3):517-525 Riccabona M, Fritz GA, Schöllnast H et al. Hydronephrotic kidney: pediatric three-dimensional US for relative renal size assessment--initial experience. Radiology. 2005;236(1):276-83. Bakker J, Olree M, Kaatee R et al. Renal volume measurements: accuracy and repeatability of US compared with that of MR imaging. Radiology 1999;211:623–628 Mancini M, Mainenti PP, Speranza A et al. Accuracy of sonografic volume measurements of kidney transplant. J Clin Ultrasound. 2006;34(4):184-9. Timberlake MD, Herndon CD. Mild to moderate postnatal hydronephrosis--grading systems and management. Nat Rev Urol. 2013;10(11):649-56. Riccabona M, Fritz G, Ring E. Potential applications of three-dimensional ultrasound in the pediatric urinary tract: pictorial demonstration based on preliminary results. Eur Radiol. 2003;13(12):2680-7. Breysem L, De Rechter S, De Keyzer F et al. 3DUS as an alternative to MRI for measuring renal volume in children with autosomal dominant polycystic kidney disease. Pediatr Nephrol. 2018;33(5):827-835. Fritz GA, Riccabona M, Bohdal G et al. Nierenvolumetrie im Kindesalter: Genauigkeit der dreidimensionalen Sonographie im Vergleich zur konventionellen Sonographie und CT / MRT [Accuracy of renal volume assessment in children by three-dimensional sonography]. Rofo. 2003 Apr;175(4):540-6. German. Brancaforte A, Serantoni S, Silva Barbosa F et al. Renal volume assessment with 3D ultrasound. Radiol Med. 2011;116(7):1095-104. Kim HC, Yang DM, Jin W et al. Relation between total renal volume and renal function: Usefulness of 3D sonographic measurements with a matrix array transducer. AJR Am J Roentgenol. 2010;194(2):W186-92. Kim HC, Yang DM, Lee SH et al. Usefulness of renal volume measurements obtained by a 3-dimensional sonographic transducer with matrix electronic arrays. J Ultrasound Med 2008;27:1673–1681 Gonçalves LF, Joshi A, Mody S et al. Volume US of the urinary tract in pediatric patients-a pilot study. Pediatr Radiol. 2011;41(8):1047-56. Cerrolaza JJ, Grisan E, Safdar N et al. Quantification of kidneys from 3D ultrasound in pediatric hydronephrosis. Annu Int Conf IEEE Eng Med Biol Soc. 2015;2015:157-60. Mory B, Somphone O, Prevost R et al. Real-time 3D image segmentation by user-constrained template deformation. Med Image Comput Comput Assist Interv. 2012;15(Pt 1):561-8. Koo TK, Li MY. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J Chiropr Med. 2016;15(2):155-63. Zou KH, Warfield SK, Baharatha A et al. Statistical validation of image segmentation quality based on a spatial overlap index. Academic Radiology. 2004;11:178–89 Shapiro SR, Wahl EF, Silberstein MJ et al. Hydronephrosis index: a new method to track patients with hydronephrosis quantitatively. Urology. 2008;72(3):536-8; discussion 538-9. Wahl EF, Lerman SE, Lahdes-Vasama TT et al. Measurement of bladder compliance can be standardized by a dimensionless number: clinical perspective. BJU Int. 2004;94(6):898-900. Rud O, Horstmann M, Aziz A et al. Prospective evaluation of intra-observer variability of the hydronephrosis index in sonographic examination of 44 patients with acute renal colic. World J Urol. 2014;32(3):691-5. Venkatesan K, Green J, Shapiro SR et al. Correlation of hydronephrosis index to society of fetal urology hydronephrosis scale. Adv Urol. 2009;2009:960490 Han JH, Song SH, Lee JS et al. Best ultrasound parameter for prediction of adverse renal function outcome after pyeloplasty. Int J Urol. 2020;27(9):775-782. Table Table 1 Underlying Kidney Diseases Pelviureteric junction stenosis n=21 Congenital hydronephrosis n=7 Primary obstructive megaureter n=7 Vesicorenal reflux n=5 Posterior urethral valve n=3 Duplex kidney n=3 Renal infundibular stenosis n=1 Urogenital Sinus n=1 Additional Declarations Competing interest reported. Mr. James R. Jago, PhD is employee of Philips Healthcare. Mrs. Laurence Rouet, PhD is employee of Philips Health Technology Innovation Paris. The study was supported by a research grant in the framework of a collaboration contract with Philips Ultrasound, Inc. Cite Share Download PDF Status: Published Journal Publication published 06 May, 2025 Read the published version in Pediatric Radiology → Version 1 posted Editorial decision: Revision requested 05 Mar, 2025 Reviews received at journal 28 Feb, 2025 Reviewers agreed at journal 16 Feb, 2025 Reviews received at journal 07 Feb, 2025 Reviews received at journal 20 Jan, 2025 Reviewers agreed at journal 19 Jan, 2025 Reviewers agreed at journal 18 Jan, 2025 Reviewers agreed at journal 07 Jan, 2025 Reviews received at journal 18 Aug, 2024 Reviewers agreed at journal 14 Aug, 2024 Reviewers agreed at journal 13 Aug, 2024 Reviewers invited by journal 13 Aug, 2024 Editor assigned by journal 12 Aug, 2024 Submission checks completed at journal 12 Aug, 2024 First submitted to journal 06 Aug, 2024 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-4868701","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":344200731,"identity":"5eac391d-3044-49a3-b3ce-bea25bee5f7c","order_by":0,"name":"Michael Esser","email":"data:image/png;base64,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","orcid":"","institution":"Universitätsklinikum Tübingen","correspondingAuthor":true,"prefix":"","firstName":"Michael","middleName":"","lastName":"Esser","suffix":""},{"id":344200732,"identity":"cf7a0411-529d-49f5-90de-d3371c35d374","order_by":1,"name":"Ilias Tsiflikas","email":"","orcid":"","institution":"Universitätsklinikum Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Ilias","middleName":"","lastName":"Tsiflikas","suffix":""},{"id":344200733,"identity":"0f69e186-8c0b-40a0-adb5-dd61e6f70f7a","order_by":2,"name":"James R. Jago","email":"","orcid":"","institution":"Philips Healthcare, Ultrasound General Imaging","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"R.","lastName":"Jago","suffix":""},{"id":344200734,"identity":"1613a196-5414-49dd-b415-1910feaab8bb","order_by":3,"name":"Laurence Rouet","email":"","orcid":"","institution":"Philips Health Technology Innovation","correspondingAuthor":false,"prefix":"","firstName":"Laurence","middleName":"","lastName":"Rouet","suffix":""},{"id":344200735,"identity":"b57327d2-e19d-4060-9412-5cada58a2178","order_by":4,"name":"Alexander Stebner","email":"","orcid":"","institution":"Universitätsklinikum Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Stebner","suffix":""},{"id":344200736,"identity":"300f5c83-60ad-487e-b93d-b05dcbbd92d2","order_by":5,"name":"Jürgen F. Schäfer","email":"","orcid":"","institution":"Universitätsklinikum Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Jürgen","middleName":"F.","lastName":"Schäfer","suffix":""}],"badges":[],"createdAt":"2024-08-06 13:06:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4868701/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4868701/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00247-025-06249-8","type":"published","date":"2025-05-06T15:57:09+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":66328765,"identity":"d67011fb-2ffa-42dc-97e2-04417324e92c","added_by":"auto","created_at":"2024-10-10 13:09:01","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":291791,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWorkflow of the study\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eImage parameters were adjusted based on the conventional 2D B-mode ultrasound. Afterwards 3D mode display showed a live side-by-side-image including a sagittal and axial view of the kidney at the same time. After 3D volume acquisition the multiplanar reformatted view shows the structure in three orthogonal planes as well as a volume-rendered image. The segmentation was performed on a separate workstation by two readers\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4868701/v1/ec9e40b51da13910411cfb5f.png"},{"id":66328766,"identity":"113506b0-b12b-4f2d-9727-f77987f518eb","added_by":"auto","created_at":"2024-10-10 13:09:02","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":54332,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBland-Altmann-Plot for all renal parenchyma values\u003c/strong\u003e\u003c/p\u003e\n\u003ch4\u003eNearly all values of renal parenchyma volume were within the limits of agreement as a tolerance range (green dashed lines). Mean difference line is depicted in red (dotted line). One outlier is below the lower limit of agreement (asterisk).\u003c/h4\u003e\n\u003cp\u003eSD: Standard deviation of the differences\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4868701/v1/b7cdb823b7b19c07d9e929f1.png"},{"id":66328724,"identity":"dbbc48eb-6148-4c14-a952-fab445e23b09","added_by":"auto","created_at":"2024-10-10 13:08:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":34268,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBoxplots of Dice similarity coefficients (DSCs) for total kidney volume, collective system and renal parenchyma\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe highest median values (bold lines) of DSC are found for total kidney volume (renal capsule; median, 0.90) and for the renal parenchyma (median, 0.88). The collective systems show lower DSCs (median, 0.77) with a relatively wide range (0.54 to 0.89). The Dice values of the renal parenchyma are lower than those of the capsule segmentation, but still at a high level. Since the segmentation of the renal capsule is semi-automatic, the high spatial agreement of the renal capsule volume seems to compensate for the discrepancies in the collective system segmentation. This can also be explained by the fact that the collective system volume is generally much smaller than the renal capsule\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4868701/v1/0fd0a663eee2f8b51fc216bd.png"},{"id":66328770,"identity":"ab552f6b-1e17-410e-bd19-cf487aefe080","added_by":"auto","created_at":"2024-10-10 13:09:03","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":23676,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation of hydronephrosis indices to hydronephrosis grading\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe lower values of hydronephrosis Index (renal parenchyma divided by total kidney capsule) going along with an increasing degree of hydronephrosis (p\u0026lt;0.001)\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4868701/v1/510b0992448104dee34bd463.png"},{"id":82537602,"identity":"72bad6cb-a858-4cc3-bf60-352092fc29d0","added_by":"auto","created_at":"2025-05-12 16:09:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1106314,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4868701/v1/3c47fcbc-eaca-4689-a620-c5cac6980686.pdf"}],"financialInterests":"Competing interest reported. Mr. James R. Jago, PhD is employee of Philips Healthcare. Mrs. Laurence Rouet, PhD is employee of Philips Health Technology Innovation Paris. The study was supported by a research grant in the framework of a collaboration contract with Philips Ultrasound, Inc.","formattedTitle":"Semiautomatic 3D Ultrasound Kidney Volume Segmentation in Pediatric Hydronephrosis: Interrater Agreement and Correlation to Conventional Hydronephrosis Grading","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn pediatric hydronephrosis the evaluation of renal parenchymal volume plays an important role for the initial estimation of prognosis and for disease monitoring. It deals as a surrogate of renal function and parenchymal damage suggesting the need for invasive diagnostic and therapeutic intervention [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The non-invasive and radiation-free sonographic assessment of renal volume is considered an indicator for the real renal size and nephron mass (renal corpuscles and canaliculi) [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Conventional ultrasound-based assessment of the kidney volume is already well established since decades; however, it is mainly based on two-dimensional (2D) measurements using the greatly simplified ellipsoid formula, which is known to underestimate real kidney volume with a high interobserver variability [\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In fact, using the ellipsoid formula can be understood as a pragmatic approach of estimating renal volume which is easy to calculate in regular non-hydronephrotic kidneys. However, bearing in mind the highly complex anatomy of the renal pelvicalyceal system, further difficulty for accurate renal volume assessment arises with hydronephrosis and possible associated malformation or status after renal surgery. In this context, the severity of pediatric hydronephrosis is a crucial point when considering the dynamic process of a growing kidney and changing degrees of renal pelvic dilation in different clinical situations [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAnteroposterior diameter with simple thresholds and calyceal dilation by themselves are insufficient parameters as they are affected by many factors like hydration status and bladder filling, and do not demonstrate the real degree of hydronephrosis in different pelvic configurations [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. The different radiology grading systems are based on subjective criteria and are thus not recommended for therapeutic decision making. As 2D ultrasound (2DUS) measurements may be inaccurate in anatomically complex kidneys, no gold standard in imaging exists for determining neither the severity of hydronephrosis nor the volume of renal parenchyma [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eToday, three-dimensional ultrasound (3DUS) has become an established imaging tool in some specialties. Since the early 2000s 3DUS of the kidney is increasingly used in the pediatric population with a reported improvement of volume assessment accuracy at a low interobserver variability [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Yet, only limited data exist on 3DUS for the evaluation of pediatric hydronephrosis where an added value might be expected compared to regular kidney morphology. Prior studies on renal 3DUS concentrated on the adult population [\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], they included non-hydronephrotic kidneys [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] and/or were restricted to preselected clinical subgroups [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn the course of the technical progress in ultrasound imaging systems, matrix array transducers have been developed for a nearly instant free-hand acquisition of volumetric data. They can be handled like conventional transducers as they do not need an electric motor or an electromagnetic position sensor. They have shown to considerably reduce volumetric errors caused by motion artifacts (which may occur in children who are not able to hold their breath or only for a short time). Initially, matrix array transducers for renal volume calculation have been evaluated in adults including a cadaver study and comparisons of normal kidney volume to CT and renal function [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. In children a matrix array transducer was used to compare renal volumes in polycystic kidneys [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. The only report on a matrix array transducer in hydronephrotic kidneys included eight kidneys for the feasibility study of a new segmentation model [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. To the best of our knowledge, no further clinical data exist on the use of a matrix array transducer in hydronephrotic kidneys in children.\u003c/p\u003e \u003cp\u003eThe purpose of our study was to evaluate the interrater agreement of kidney volume segmentation by three-dimensional ultrasound (3DUS) with a matrix array transducer in children with hydronephrosis and compare the resulting metrics to conventional hydronephrosis grading.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e This prospective monocenter study was approved by the local ethics committee. Signed informed consent was obtained from all legal guardians and children above the age of 11 participating in the study. All study procedures were conducted in accordance with the \u003cem\u003eGuidelines for Good Clinical Practice\u003c/em\u003e and ethical standards as laid down in the 1964 \u003cem\u003eDeclaration of Helsinki\u003c/em\u003e and its later amendments or comparable ethical standards. The authors had full control of the data.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy population\u003c/h2\u003e \u003cp\u003eBetween March and September 2021 all patients under 17 years who were referred to our sonography center as part of the in-house pediatric urology consultation with a clinical indication for ultrasound of the kidneys and present hydronephrosis were included in the study. Patients were excluded from the study if patients or legal guardians declined afterwards to participate in the study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e3D Ultrasound Acquisition and Evaluation\u003c/h2\u003e \u003cp\u003eInitially, routine 2DUS was performed by a radiologist with 7 years of experience in ultrasound using an EPIQ 5G ultrasound device (Philips Medical Systems, Bothell, WA, USA). The maximum anteroposterior diameter of renal pelvis was measured on transverse 2D B-mode images and hydronephroses were classified grade 1\u0026ndash;4 following the Onen-2007 hydronephrosis grading system [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAfterwards, freehand 3DUS with a linear matrix array transducer (X6-1) was performed by the same radiologist immediately after 2D sonography (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). For this purpose, all patients were placed in prone position. First, image parameters were adjusted with the help of protocols which were optimized for children by the manufacturer and stored on the device. With activation of the 3D X-plane mode, a live side-by-side-image included simultaneous sagittal and axial views of the kidney to ensure that the entire kidney will be included in the 3D volume. The acquisition scan of one single data set for each kidney took two to three seconds. The multiplanar reformatted (MPR) display showed the structure in three orthogonal planes as well as a volume-rendered image to check the volume for complete coverage of the kidney. For each kidney three subsequent volumetric data sets were acquired. The 3D data sets were automatically stored on an integrated hard disc. Time for the acquisition of the three data sets was between one and three minutes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSemi-automated 3D kidney segmentation was performed on a separate workstation using research prototype software (Philips Health Technology Innovation, Paris, France). The prototype segmentation is based on implicit shape deformation [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The principle is to deform a 3D kidney template shape according to internal and external forces. The forces are governed by fixed parameters tuned for 3D Ultrasound of kidney volumes. After initial deformation, the reader can apply interactive corrections to adjust the segmentation to the exact kidney capsule borders. In a second step, a region growing algorithm is used to interactively segment the collective system. The collective system segmentation may be composed of multiple sub-volumes. Upon completion, segmentations and corresponding volumes quantifications are saved. Two readers segmented the kidneys freely choosing one of the three saved data sets for each kidney. The readers were instructed to choose a data set with respect to the best coverage of organ surface and with minimal motion artifacts. The readers were blinded to all identifying and clinical data.\u003c/p\u003e \u003cp\u003eFor renal capsule segmentation a sphere was placed in an axial view to define the center of the kidney. The sphere was used to initialize the size and pose of the kidney 3DUS template shape, which was then deformed. Afterwards, the ellipsoid capsule was adjusted by providing additional marking points in the upper and lower poles of the kidney using a sagittal view. In a separate edition mode, the intra- and extrarenal portions of the (dilated) collective system including calices were segmented. The resulting segmentation volumes were checked in all three orthogonal planes. Quantitative results were shown in a separate table. The following values were calculated: collective system, renal parenchyma, total kidney volume (addition of collective system and renal parenchyma), hydronephrosis index (HI, kidney parenchyma divided by total renal capsule). The extrarenal proportion of the pelvis was not included for the calculation of HI.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eStatistics\u003c/h2\u003e \u003cp\u003eFor statistical analysis, IBM SPSS Statistics (version 22 for Windows, Ehningen, Germany) was used. The Kolmogorow-Smirnov test was used to study the distribution of quantification data. Continuous variable data are presented as means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviations. Data that did not follow a normal distribution are presented as median with interquartile ranges (IQR).\u003c/p\u003e \u003cp\u003eBland\u0026ndash;Altman plots were used to visually compare measurements of the two readers for the kidney parenchyma. The mean difference and the upper and lower limits of agreement (LoA) were calculated. The limits of agreement were defined as the mean difference\u0026thinsp;\u0026plusmn;\u0026thinsp;1.96 standard deviations of the differences.\u003c/p\u003e \u003cp\u003eTo assess the interrater agreement, intraclass correlation coefficient (ICC) estimates and their 95% confidence levels (CL) were calculated using an absolute agreement, two-way mixed-effects model [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The ICC was defined as slight (0\u0026ndash;0.20), fair (0.21\u0026ndash;0.40), moderate (0.41\u0026ndash;0.60), substantial (0.61\u0026ndash;0.80), and excellent (0.81\u0026ndash;1.00). Spearman's rank correlation test was used to analyze the relationship between hydronephrosis grading and the collective system volumes. A p-value less than 0.05 was considered to indicate statistical significance.\u003c/p\u003e \u003cp\u003eTo evaluate the voxel-wise spatial overlap of the segmentations, Dice similarity coefficients (DSC) were computed between the 22 segmentations of the two readers in which both readers selected the same data set [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. The values of DSC range from 0, representing no spatial overlap between two sets of segmentations, to 1, indicating a complete overlap.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003ePatient characteristics\u003c/h2\u003e \u003cp\u003e48 kidney volumes were calculated in 45 children (median age 4.5 years; range, 1 month to 16 years; interquartile range, IQR, 7 years; 7 patients under the age of 12 months; 23 patients under the age of 5 years; 35 males, 10 females). The left side (n\u0026thinsp;=\u0026thinsp;29, 60%) was a little more frequently affected than the right side (n\u0026thinsp;=\u0026thinsp;19, 40%).\u003c/p\u003e \u003cp\u003eThe most frequent hydronephrosis grading was grade 2 (n\u0026thinsp;=\u0026thinsp;29), followed by grade 3 (n\u0026thinsp;=\u0026thinsp;15). Only 4 patients were included with grade 1 hydronephrosis. The underlying kidney diseases of all patients are summarized in Table\u0026nbsp;1. In 18/48 kidneys (38%) a prior history of urological surgery was recorded.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eInterrater Agreement\u003c/h2\u003e \u003cp\u003eThe two readers chose to use the same data set for segmentation in 46% (22/48) of the cases. For an assessment of the range of all kidney parenchyma values the Bland-Altmann-Plot is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The mean difference was 0.68 ml. The upper limit of agreement was 25.3 ml and the lower limit of agreement was \u0026minus;\u0026thinsp;24.0 ml. One outlier was identified below the lower limit of agreement: In this 11-year old patient with grade 3 hydronephrosis due to renal infundibular stenosis, parenchyma value difference between the two readers was 65 ml. The collective system was segmented with similar volumes of 39.6 and 40.4. ml, but a relevant discrepancy occurred in renal capsule segmentation (111 vs. 177 ml). In this case, different data sets were used for kidney segmentation. The subsequent manual comparison showed that the upper kidney pole was not fully captured in the reader 1 data set.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eInterrater agreement between the two readers for total kidney volume (renal capsule; ICC 0.94; 95% CL [0.90; 0.97]; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), collective system (ICC 0.87; 95% confidence level (CL) [0.78; 0.93]; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), HI (ICC 0.83; 95% CL [0.71; 0.90]; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and renal parenchyma (ICC 0.92; 95% CL [0.86; 0.96]; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) was good to excellent. Renal parenchyma values had a median of 62.6 and 62.0 ml for the two readers, ranging from 16.1 ml to 147.7 ml (reader 1; IQR, 37.9 ml) and from 16.5 ml to 139.5 ml (reader 2; IQR 41.3 ml). The HI had a median of 0.91 for both readers, ranging from 0.63 to 0.99 (reader 1; IQR, 0.10) and from 0.52 to 0.98 (reader 2; IQR 0.12).\u003c/p\u003e \u003cp\u003eThe 22 segmentations with the same data set achieved a median DSC of 0.90 ranging from 0.76 to 0.95 (IQR, 0.06) for the total kidney volume (renal capsule), a median DSC of 0.77 ranging from 0.54 to 0.89 (IQR, 0.76) for the collective system and a median DSC of 0.88 ranging from 0.73 to 0.94 (IQR, 0.10) for the renal parenchyma. The corresponding boxplots for the three DSCs between reader 1 and reader 2 by segmentation pairs are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. In one case a remarkably low DSC for the collective system occurred (DSC, 0.54). In this 8-year old patient with duplex kidney, similar absolute values were obtained for the volumes of the slightly dilated renal pelvis (grade 1; 1.4 vs. 1.5 ml), although the spatial overlap was obviously not optimal. This also resulted in a similar renal volume with a difference between the readers of 4.7 ml (77.2 vs. 72.6 ml).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eHydronephrosis Grading\u003c/h2\u003e \u003cp\u003eHydronephrosis indices (HI) of both readers were 0.91 (median; IQR, 0.1 for both readers; range for reader 1, 0.63 to 0.99; range for reader 2, 0.52 to 0.98).\u003c/p\u003e \u003cp\u003eThere was a positive correlation between the hydronephrosis grading and the HI (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) and between hydronephrosis grading and the volume of the collective system (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study provides initial data concerning the feasibility of a free-hand three-dimensional kidney volume assessment for the evaluation of hydronephrosis in children. The workflow presented here allowed for reproducible and accurate results of volumetric data providing useful information according to established methods of hydronephrosis grading. Measurements can be obtained rapidly and easily, without the use of radiation, contrast agent or anesthesia. The method is thus feasible in neonates, infants and children. The prototype software for semiautomatic segmentation as one component of the presented workflow (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) allows a reliable subtraction of the dilated collecting system and enables reproducible calculation of renal parenchyma size.\u003c/p\u003e \u003cp\u003eA recent review about modern approaches of kidney imaging in children mentioned 3DUS as the \u0026ldquo;least invasive and most cost-effective way\u0026rdquo; for repetitive imaging with monitoring of renal size in children and underlined the importance of upcoming investigation on 3DUS as an \u0026ldquo;important goal for future research\u0026rdquo; [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. As part of this development the first clinical studies reported the use of 3DUS as a promising tool with superior accuracy for the depiction of renal anatomy and reliable parenchyma volume determination in children with renal malformations and for postoperative assessment compared to planimetric analysis [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. In our study, a relevant portion of the cohort had congenital renal abnormalities (Table\u0026nbsp;1) and/or complex anatomy of the urinary tract, like urogenital sinus or duplex kidney. Renal surgery was another influencing factor that altered renal anatomy in some patients (n\u0026thinsp;=\u0026thinsp;18, 38%). Therefore, our results meet the former expectations and confirm the favored application of matrix array 3DUS as a robust method in anatomically complex situations.\u003c/p\u003e \u003cp\u003eA further study recommended to prefer 3DUS over MRI for accurate kidney volume assessment in the follow-up of children with autosomal dominant polycystic kidney disease (ADPKD) [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In contrast to our study, no hydronephrotic kidneys were evaluated and only older patients (\u0026gt;\u0026thinsp;8 years) were included. In our study, a relevant proportion of the patients was markedly younger (51% under the age of 5 years) including infants (under the age of 12 months; n\u0026thinsp;=\u0026thinsp;7) bearing the risk of movement during the 3D volume acquisition. Thus, the fast acquisition by means of a matrix array transducer can assure a reliable volume assessment in a representative patient cohort concerning age distribution.\u003c/p\u003e \u003cp\u003eIn the comparative study of Fritz et al. infants under the age of 6 months were included, however, additional six infants were excluded from the analysis because of increased motion resulting in unusable image data [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Riccabona et al. also reported that 3DUS was not possible in some uncooperative infants that were excluded from the study [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In our prospective study, no patients were excluded afterwards and at least one of the three acquired data sets was evaluable by the software in all patients.\u003c/p\u003e \u003cp\u003eRiccabona et al. assessed hydronephrotic kidneys with an electromagnetic positioning device and a mechanically driven transducer, but did not use an electronic matrix array transducer. [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In the similar study of Fritz et al. only eight patients with hydronephrosis were analyzed [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Both prior studies [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] already used semiautomated volume calculation with threshold-based segmentation. In our study a comparable semiautomated segmentation model was used to minimize time requirements. There are some methodological aspects to consider with regard to the semiautomatic segmentation used: Our evaluation has shown that the reliable and exact detection of the kidney boundaries is particularly important for accurate parenchyma assessment. Since the volume of the renal capsule is much larger than that of the renal pelvis, inaccurate measurements of the renal pelvis can also be compensated for. This explains why one case with a large volume deviation of the renal capsule between the readers was an outlier in the comparison of the parenchyma values (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The selection of the data set to be segmented could also be important here: In the case mentioned above, two different data sets were segmented. We consider it useful that in routine use with moving children (without the ability to stop breathing), selection from several 3D data can allow greater certainty for accurate subsequent segmentation. However, it only seems to be effective if the renal borders, in particular the upper and lower poles, have been completely captured in all data sets. The time required to save multiple data sets is minimal with the matrix transducer.\u003c/p\u003e \u003cp\u003eThe initial studies that evaluated matrix array transducers for renal volume calculation in adults compared normal kidney volume to CT [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] and investigated correlations between renal volume and renal function parameters, but did not evaluate the reproducibility of the technique [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Both studies valued 3DUS with a matrix array transducer as a reliable tool for determining renal volumes with reduced measurement errors and recommended to study the reproducibility in different clinical cohorts [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The only report on a matrix array transducer in hydronephrotic kidneys included eight kidneys for the feasibility study of a new segmentation model [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. In this study the renal volumes were analyzed with respect to segmentation errors and relative volume differences, but also no interrater variability was evaluated.\u003c/p\u003e \u003cp\u003eIn a prior study the hydronephrosis index (HI) has been proposed as a dimensionless parameter of quantifying hydronephrosis in children by 2DUS [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. The advantage of standardization in using dimensionless metrics in medical imaging has been mentioned previously [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Rud et al. reported an excellent interobserver agreement for HI and a correlation between HI and the sonographic degree of renal pelvis dilation in adults with stone-related renal colic [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. In a prior study, a close correlation between HI and the sonographically evaluated grade of hydronephrosis was shown in patients with pelvi-ureteric junction obstruction [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. These study conclusions are in line with our own results with a good interrater agreement (ICC 0.83) and a narrow correlation between HI and the hydronephrosis grading (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Han et al. found that preoperative HI in patients for pyeloplasty can work as a possible prognostic marker for adverse renal function outcome [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. The clinical value of 3D ultrasound parameters has not been evaluated in our study. For further studies on HI as an indicator of pediatric renal function in everyday clinical use it has to be kept in mind that 2D sonographic calculation of the index may be time-consuming [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Our results show that semiautomatic calculation of HI in 3D data sets provides a simplified way of objective measurement which is recommendable for further investigations.\u003c/p\u003e \u003cp\u003eOur study has the following limitations: In contrast to prior studies we did not perform any comparisons to other imaging modalities (CT or MRI) or to 2DUS. However, the general feasibility and advantages of 3D sonographic evaluations over conventional ultrasound measurements have already been published. A technical limitation is that for all children a transducer with a frequency range of 6\u0026thinsp;\u0026minus;\u0026thinsp;1 MHz was used, which resulted in low image resolution in infants and neonates. However, complete coverage of the renal capsule without interpolation of the upper and/or lower pole contour was thus easily possible and image contrast was estimated as suitable for the differentiation of the collective system in all patients. No grade 4 hydronephroses were included, which reduces the generalizability of the results.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eNovel semiautomatic 3DUS volumetric analysis has a high degree of interrater agreement providing parenchyma volume in hydronephrotic kidneys. Volumes of the collective system and hydronephrosis index correlate with the extent of hydronephrosis.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMr. James R. Jago, PhD is employee of Philips Healthcare. Mrs. Laurence Rouet, PhD is employee of Philips Health Technology Innovation Paris. The study was supported by a research grant in the framework of a collaboration contract with Philips Ultrasound, Inc.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contribution\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eM.E.: Conceptualization, Formal analysis and investigation, Original draft preparation, Funding acquisition, SupervisionI.T.: Methodology, Formal analysis and investigation, SupervisionJ.R.J.: Conceptualization, Funding acquisition, ResourcesL.R.: Methodology, Funding acquisition, Resources, Original draft preparationA.S.: Formal analysis and investigation, Original draft preparationJ.F.S.: Conceptualization, Methodology, Funding acquisition, Resources, SupervisionAll authors reviewed the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available from Philips Healthcare. Restrictions apply to the availability of these data, which were used under license for this study.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eOnen A. Grading of Hydronephrosis: An Ongoing Challenge. Front Pediatr. 2020;8:458.\u003c/li\u003e\n \u003cli\u003eDeFreitas MJ, Katsoufis CP, Infante JC et al. The old becomes new: advances in imaging techniques to assess nephron mass in children. Pediatr Nephrol. 2021;36(3):517-525\u003c/li\u003e\n \u003cli\u003eRiccabona M, Fritz GA, Sch\u0026ouml;llnast H et al. Hydronephrotic kidney: pediatric three-dimensional US for relative renal size assessment--initial experience. Radiology. 2005;236(1):276-83.\u003c/li\u003e\n \u003cli\u003eBakker J, Olree M, Kaatee R et al. Renal volume measurements: accuracy and repeatability of US compared with that of MR imaging. Radiology 1999;211:623\u0026ndash;628\u003c/li\u003e\n \u003cli\u003eMancini M, Mainenti PP, Speranza A et al. Accuracy of sonografic volume measurements of kidney transplant. J Clin Ultrasound. 2006;34(4):184-9.\u003c/li\u003e\n \u003cli\u003eTimberlake MD, Herndon CD. Mild to moderate postnatal hydronephrosis--grading systems and management. Nat Rev Urol. 2013;10(11):649-56.\u003c/li\u003e\n \u003cli\u003eRiccabona M, Fritz G, Ring E. Potential applications of three-dimensional ultrasound in the pediatric urinary tract: pictorial demonstration based on preliminary results. Eur Radiol. 2003;13(12):2680-7.\u003c/li\u003e\n \u003cli\u003eBreysem L, De Rechter S, De Keyzer F et al. 3DUS as an alternative to MRI for measuring renal volume in children with autosomal dominant polycystic kidney disease. Pediatr Nephrol. 2018;33(5):827-835.\u003c/li\u003e\n \u003cli\u003eFritz GA, Riccabona M, Bohdal G et al. Nierenvolumetrie im Kindesalter: Genauigkeit der dreidimensionalen Sonographie im Vergleich zur konventionellen Sonographie und CT / MRT [Accuracy of renal volume assessment in children by three-dimensional sonography]. Rofo. 2003 Apr;175(4):540-6. German.\u003c/li\u003e\n \u003cli\u003eBrancaforte A, Serantoni S, Silva Barbosa F et al. Renal volume assessment with 3D ultrasound. Radiol Med. 2011;116(7):1095-104.\u003c/li\u003e\n \u003cli\u003eKim HC, Yang DM, Jin W et al. Relation between total renal volume and renal function: Usefulness of 3D sonographic measurements with a matrix array transducer. AJR Am J Roentgenol. 2010;194(2):W186-92.\u003c/li\u003e\n \u003cli\u003eKim HC, Yang DM, Lee SH et al. Usefulness of renal volume measurements obtained by a 3-dimensional sonographic transducer with matrix electronic arrays. J Ultrasound Med 2008;27:1673\u0026ndash;1681\u003c/li\u003e\n \u003cli\u003eGon\u0026ccedil;alves LF, Joshi A, Mody S et al. Volume US of the urinary tract in pediatric patients-a pilot study. Pediatr Radiol. 2011;41(8):1047-56.\u003c/li\u003e\n \u003cli\u003eCerrolaza JJ, Grisan E, Safdar N et al. Quantification of kidneys from 3D ultrasound in pediatric hydronephrosis. Annu Int Conf IEEE Eng Med Biol Soc. 2015;2015:157-60.\u003c/li\u003e\n \u003cli\u003eMory B, Somphone O, Prevost R et al. Real-time 3D image segmentation by user-constrained template deformation. Med Image Comput Comput Assist Interv. 2012;15(Pt 1):561-8.\u003c/li\u003e\n \u003cli\u003eKoo TK, Li MY. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J Chiropr Med. 2016;15(2):155-63.\u003c/li\u003e\n \u003cli\u003eZou KH, Warfield SK, Baharatha A et al. Statistical validation of image segmentation quality based on a spatial overlap index. Academic Radiology. 2004;11:178\u0026ndash;89\u003c/li\u003e\n \u003cli\u003eShapiro SR, Wahl EF, Silberstein MJ et al. Hydronephrosis index: a new method to track patients with hydronephrosis quantitatively. Urology. 2008;72(3):536-8; discussion 538-9.\u003c/li\u003e\n \u003cli\u003eWahl EF, Lerman SE, Lahdes-Vasama TT et al. Measurement of bladder compliance can be standardized by a dimensionless number: clinical perspective. BJU Int. 2004;94(6):898-900.\u003c/li\u003e\n \u003cli\u003eRud O, Horstmann M, Aziz A et al. Prospective evaluation of intra-observer variability of the hydronephrosis index in sonographic examination of 44 patients with acute renal colic. World J Urol. 2014;32(3):691-5.\u003c/li\u003e\n \u003cli\u003eVenkatesan K, Green J, Shapiro SR et al. Correlation of hydronephrosis index to society of fetal urology hydronephrosis scale. Adv Urol. 2009;2009:960490\u003c/li\u003e\n \u003cli\u003eHan JH, Song SH, Lee JS et al. Best ultrasound parameter for prediction of adverse renal function outcome after pyeloplasty. Int J Urol. 2020;27(9):775-782.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"461\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e Underlying Kidney Diseases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003ePelviureteric junction stenosis\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=21\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003eCongenital hydronephrosis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=7\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003ePrimary obstructive megaureter\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=7\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003eVesicorenal reflux\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003ePosterior urethral valve\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003eDuplex kidney\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003eRenal infundibular stenosis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"85.68329718004338%\"\u003e\n \u003cp\u003eUrogenital Sinus\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.316702819956616%\"\u003e\n \u003cp\u003en=1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\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":"pediatric-radiology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"prad","sideBox":"Learn more about [Pediatric Radiology](http://link.springer.com/journal/247)","snPcode":"247","submissionUrl":"https://submission.nature.com/new-submission/247/3","title":"Pediatric Radiology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Hydronephrosis, Pediatrics, Kidney Diseases, Imaging, Three-Dimensional, Ultrasonography","lastPublishedDoi":"10.21203/rs.3.rs-4868701/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4868701/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Two-dimensional assessment of the kidney volume underestimates real kidney volume with a high interobserver variability. Limited data exist on innovative 3D ultrasound (3DUS) technique for the evaluation of pediatric hydronephrosis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eObjective: \u003c/strong\u003eTo evaluate the interrater agreement of kidney volume segmentation by 3DUS with a matrix array transducer in children with hydronephrosis and comparison of 3D metrics to conventional hydronephrosis grading.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMaterials and methods:\u003c/strong\u003e 48 kidney volumes were prospectively acquired in 45 patients with hydronephrosis by freehand 3DUS (6-1MHz volumetric sector array, electronic rotation; median age, 4years; 1month to 16years). Semi-automated kidney segmentation was performed by two independent readers providing volumes for total kidney (renal capsule), dilated collective system, renal parenchyma (renal capsule - collective system) and hydronephrosis index (renal parenchyma / renal capsule). Interrater agreement was evaluated with Bland–Altman plots, intraclass correlation coefficient (ICC) and Dice similarity coefficients. The maximum 2D diameter of renal pelvis was measured and hydronephroses were morphologically classified grade 1-4.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e Interrater agreement for renal capsule, collective system, hydronephrosis index and renal parenchyma was good to excellent with ICC of 0.94, 0.87, 0.83 and 0.92 respectively (p\u0026lt;0.001 each). Median Dice was 0.90 (capsule), 0.77 (collective system) and 0.88 (parenchyma). There was a positive correlation between hydronephrosis grading and ultrasonic hydronephrosis index and between renal pelvis diameter and collective system volume (p\u0026lt;0.001 both).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion:\u003c/strong\u003e Semiautomatic 3DUS volumetric analysis has a high degree of interrater agreement providing parenchyma volume in hydronephrotic kidneys. Volumes of the collective system and hydronephrosis index correlate with the extent of hydronephrosis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrial registry:\u003c/strong\u003e trial registration number, DRKS00022772; date of registration, 07/31/2020\u003c/p\u003e","manuscriptTitle":"Semiautomatic 3D Ultrasound Kidney Volume Segmentation in Pediatric Hydronephrosis: Interrater Agreement and Correlation to Conventional Hydronephrosis Grading","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-10 13:08:21","doi":"10.21203/rs.3.rs-4868701/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-03-05T06:36:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-28T15:14:19+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"277356587641219468696727020422852873073","date":"2025-02-16T21:49:12+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-07T16:57:42+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-01-20T05:19:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"45291099113159108067042897126732117928","date":"2025-01-19T14:10:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"233764646462017276348740635427660947686","date":"2025-01-18T19:41:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"184286626553426809116965363903741376291","date":"2025-01-07T23:18:41+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-08-18T10:17:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"44158364704644238290809391613608807947","date":"2024-08-14T12:59:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"234570341169918931292690665636909076667","date":"2024-08-14T03:46:10+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-08-13T09:52:19+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-08-13T03:34:20+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-08-13T03:33:49+00:00","index":"","fulltext":""},{"type":"submitted","content":"Pediatric Radiology","date":"2024-08-06T13:05:09+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"pediatric-radiology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"prad","sideBox":"Learn more about [Pediatric Radiology](http://link.springer.com/journal/247)","snPcode":"247","submissionUrl":"https://submission.nature.com/new-submission/247/3","title":"Pediatric Radiology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"3225da39-9ff6-4af3-b8f7-b2e6a8da5379","owner":[],"postedDate":"October 10th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-05-12T16:04:05+00:00","versionOfRecord":{"articleIdentity":"rs-4868701","link":"https://doi.org/10.1007/s00247-025-06249-8","journal":{"identity":"pediatric-radiology","isVorOnly":false,"title":"Pediatric Radiology"},"publishedOn":"2025-05-06 15:57:09","publishedOnDateReadable":"May 6th, 2025"},"versionCreatedAt":"2024-10-10 13:08:21","video":"","vorDoi":"10.1007/s00247-025-06249-8","vorDoiUrl":"https://doi.org/10.1007/s00247-025-06249-8","workflowStages":[]},"version":"v1","identity":"rs-4868701","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4868701","identity":"rs-4868701","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}