Automated determination of bone age and bone health index in pediatric liver transplant recipients

Automated determination of bone age and bone health index in pediatric liver transplant recipients | 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 Automated determination of bone age and bone health index in pediatric liver transplant recipients Gerardo Napodano, Fabian Kück, Quirin Strotzer, Birgit Knoppke, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8694835/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Background: Bone metabolism in children who have undergone pediatric liver transplantation (pLT) can be negatively affected, particularly in the presence of cholestasis. Pediatric bone status is a critical determinant of bone health through adulthood. The aim of this study was to evaluate bone age (BA), a marker of skeletal maturity, and bone health index (BHI), a surrogate marker of bone density, in pLT recipients. Methods: A total of 449 left hand radiographs of 207 patients [median age at LT: 1.1 years (IQR: 0.43; 6.8 years); female = 50.2%], all entered in the local pLT register, were evaluated in this retrospective, IRB-approved single-center exploratory study. Fully automated determination of BA-standard deviation score (BA SDS) and BHI SDS was performed using a commercial, CE-marked AI tool and relationships with age, sex, underlying disease (biliary atresia vs. non-biliary atresia), height percentile and plasma parathyroid hormone (PTH) were assessed using t-tests, permutation tests, and Pearson correlation in both the whole cohort and the subgroup with biliary complications. Results: Mean BA SDS pre-LT was 0.03 ± 1.50, 0.03 ± 1.70 at year 1, -0.07 ± 1.62 at year 3 and − 0.21 ± 1.63 at year 5. Mean BHI SDS pre-LT was − 1.8 ± 1.20, -1.6 ± 1.20 at year 1, -2.0 ± 1.10 year 3 and − 2.2 ± 1.3 at year 5. A significant positive correlation was observed between BA SDS and height percentiles at all timepoints. PTH showed a significant inverse correlation with BA SDS at years 3 and 5 and with BHI SDS at year 5. No significant differences in mean BA SDS or BHI SDS were observed between groups with or without biliary complications or PTCD/ERCP at any timepoint. Conclusions: While mean BA SDS remained stable around zero post-LT, mean BHI SDS was consistently reduced and declined progressively until year 5, indicating decreased bone mineral density. Further prospective studies with larger cohorts are needed to determine the utility of automated BA and BHI assessment following pLT. Child/adolescent liver transplantation bone age measurement radiography hand bone density Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Pediatric liver transplantation (pLT) is the gold standard for life-threatening acute and chronic liver diseases and multiple liver-based inherited metabolic defects and should be considered for all children with end-stage liver disease 1 , 2 . Since long-term patient and graft survival are excellent, attainment of normal growth and development are crucial to support children in leading a life that is as close to normal as possible and to be an independent adult. This is particularly relevant for very young pLT recipients, where infantile-onset liver disease and therapeutic measures, in particular surgery and immunosuppression, can adversely affect growth and neurodevelopment 3 . Pediatric bone status entering adulthood is considered a critical determinant of bone health through adulthood. Bone status may be adversely affected due to pre-existing bone disease, effects of medications and nutritional challenges related to pLT 4 . Elevated fracture risk in patients following liver and other solid organ transplantation has been described in children as well as adults 5 – 7 . Determination of bone age (BA) to assess skeletal maturity and predict adult height is applied in various pediatric diseases and their associated therapeutic interventions, including chronic liver diseases 8 , 9 . Since manual BA determination is time-consuming and associated with considerable inter-rater variability 10 , fully automated image analysis methods have been developed. The method uses a validated machine-learning approach for BA prediction. It has been shown to outperform single manual readings and to produce fewer major errors, indicating that it may be safer than manual assessment 11 12 . Another promising biomarker of skeletal health in children is the Bone Health Index (BHI) that is based on the automated assessment of metacarpal thickness, width, length and medullary diameter 13 , 14 . At our institution, the first pLTs were performed in 2008 and have been ongoing ever since. Normal patient follow-up care takes place at the outpatient LT clinic and data obtained here is recorded in the pLT register. Left hand radiographs have been taken during annual outpatient visits to determine BA and its progression. Given the absence of systematic data on automated BA and BHI assessments in pLT patients, our study examined whether BA SDS and BHI SDS deviate from normative standards in this population. Material and Methods Study design and patient population This retrospective single-center study was performed in accordance with the ethical principles of the Declaration of Helsinki and its amendments and with the approval of the local IRB (vote number: 24-4012-104). The IRB waived the requirement for written consent. The local picture archiving and communication system (PACS) was queried for hand radiographs (a.p. or d.v.) of patients < 18 years, acquired between January 1, 2008 and April 30, 2024, and whose reports included the search terms ‘liver transplant’, ‘LT’, or ‘bone age’. This yielded 324 patients, of whom 31 were excluded because no liver transplant had been performed or no post-transplant radiograph was available. The data for the remaining 293 individuals were compared with the entries in the local pLT-registry that is based on the password-encrypted QNOME study platform. 50 individuals were not registered due to various reasons (i.e., pLT in patients who had turned 18, loss to follow-up) and were therefore excluded. 36 patients were ineligible due to unsuitable timing of the X-ray examination. (Fig. 1 ). For the remaining 207 patients who met the inclusion criteria, descriptive data were extracted from the registry, local PACS and hospital information system. Data collection of pre-transplant demographics included underlying disease, age and sex of recipient at LT; post-transplant details explored included biliary complications, PTCD and/or ERCP-therapy (including bile duct stenting), and fractures. Various types of complications are subsumed in the register under the umbrella term „biliary complications“: intrahepatic duct stenosis (at the anastomosis or extrahepatic); biliary leckage/bilioma; cholangitis; ischemic type biliary lesion; biliary cast/sludge; secondary sclerosing cholangitis; necrosis; others. Height percentile and plasma PTH at different timepoints (preparation for surgery and 1-, 3-, and 5-years post LT) were noted. Automated bone age estimation BoneXpert has been used for automated, PACS-integrated bone age assessment at our institution since May 2023. All images relevant to the study that were taken before this date were retrospectively evaluated using a standalone solution tailored for research purposes. BoneXpert automatically calculates bone age according to the Greulich and Pyle and Tanner Whitehouse standards in a process that takes less than 15 s per radiograph via a Digital Imaging and Communications in Medicine (DICOM) node. No data are stored, shared or transferred outside the local PACS. Chronological age, BA according to Greulich and Pyle (GP), the most common used bone age estimation technique 15 , and BA SDS, BHI and BHI SDS were documented for each radiograph (Fig. 2 ). Bone Age. BA SDS > 0 indicates advanced bone age, BA SDS < 0 indicates delayed bone age. The overall reported BA (GP) accuracy is 0.62 years, with a root mean square error and a mean absolute deviation of 0.30 and 0.21 years in infants, respectively 16 . The underlying mathematical framework has been described previously 17 . An extension of the method at the end of puberty, up to a BA of 19 years for boys and 18 years for girls of Caucasian (most prevalent ethnicity in this study cohort) origin has been reported 18 . Bone Health Index. Accuracy is defined as the degree of agreement with a reference method. Since BHI represents a concept derived by automated radiogrammetry, no such external reference exists. Consequently, BHI values are expressed in arbitrary units. Interpretation relies on comparison with a reference curve, which serves as the internal reference standard. As this curve is obtained using the same device, BHI accuracy relative to the reference equals the method’s precision, characterized by a relative SD of 1.4% at a cortical thickness of 1.3 mm (appr. 10 years of age) 19 (Fig. 3 ). The underlying parameters and methodological analysis have been discussed elsewhere 13 , 19 . Statistical analysis Variables were summarized using absolute and relative frequencies, mean ± standard deviation, median, range and interquartile range (IQR). For the comparison of different groups, we applied Fisher’s exact test for nominal variables, Welch’s t test for BA SDS and BHI SDS, and the non-parametric Studentized permutation test by Neubert and Brunner for all other numeric variables 20 . Moreover, we computed Pearson correlation coefficients and fitted multivariable linear regression models. Results are reported with 95% confidence intervals (CI) and visualized using estimated marginal means. The significance level was set to α= 5% for all statistical tests. Due to the exploratory nature of this study, no adjustment for multiple testing was applied. All analyses were performed with the statistical programming environment R (version 4.2.3; R Core Team (2023)) using the R packages nparcomp (version 3.0) 21 for the studentized permutation test, and ggstatsplot (version 0.12.0) 22 and ggeffects (version 2.3.1) 23 for the visualizations. Results 106 patients (51.2%) presented with biliary atresia, 101 (48.8%) with other diagnoses, the most common of which being non A-E hepatitis (6.8%), progressive familial intrahepatic cholestasis (4.8%), biliary hypoplasia in Alagille`s disease (3.9%) and cystic fibrosis (3.4%). 104 patients (50.2%) were female, 103 male (49.8%). Median age at LT was 1.1 years (IQR: 0.43; 6.8 years). Mean age at the time of LT was 4.1 ± 5.2 years. The standard deviation is attributable to the fact that in this group, many children are very young, but some significantly older children are included, explaining the high dispersion. 83% of the recipients had received a split liver graft, 17% a whole organ. 23 patients died between 2008 and 2024. Standard immunosuppressive therapy included corticosteroid administration. All pLT patients underwent serum level–guided vitamin D supplementation. Determination of BA and its progression was advised in patients pre-LT except for neonates, and post-LT at the annual outpatient visits except for patients in whom epiphyseal closure had already occurred. 449 radiographs were evaluated. 13 images were automatically rejected. For the rejected images, BHI could not be determined, and BA was evaluated manually. There were 37 images prior to LT, 167 at year 1, 128 at year 3 and 117 at year 5. BA SDS and BHI SDS Mean BA SDS remained around the zero value at all time points in the overall cohort (LT: 0.03 ± 1.50; 1 year: 0.03 ± 1.70; 3 years: −0.07 ± 1.62; 5 years: −0.21 ± 1.63; Fig. 4 a). The difference from LT to year 1 was significant (mean − 0.57 ± 1.22, p = 0.022) in the subset with available paired measurements; however, patients with an available BA measurement at LT had lower BA SDS at one year than those without a baseline measurement. As with BA, patients with an available BHI measurement at the time of LT had lower BHI at 1 year than those without a baseline measurement. The difference between year 1 and 3 (mean − 0.01 ± 0.97, p = 0.310) and between 3 and 5 (mean − 0.13 ± 0.80, p = 0.117) was not significant. Mean BHI SDS at the time of LT was − 1.8 ± 1.20, -1.6 ± 1.20 one year post LT, -2.0 ± 1.10 at the third-year check-up and − 2.2 ± 1.3 five years post LT (Fig. 4 b). The change in BHI SDS between the 1st and 3rd year (mean − 0.40 ± 1.10) as well as between 3rd and 5th year (mean − 0.16 ± 0.71) was significant with a p-value < 0.001 and 0.032, respectively. As with BA, patients with a BHI measurement at LT had lower mean BHI SDS at 1 year than those without a baseline measurement; however, the difference from LT to year 1 (mean − 0.19 ± 0.73) was not statistically significant in the subset with available paired measurements (p = 0.204) with a relatively wide CI reflecting the limited subgroup size (n = 26, 95% CI [-0.48; 0.11]). BA SDS and BHI SDS were not found to be significantly influenced by patients' sex. Biliary atresia In the 106 patients with biliary atresia, the proportion of females (54.7%) was greater than that of males (45.3%) compared to patients with other underlying diseases (45.5% female and 54.5% male) with no statistically significant difference (p = 0.212). Mean age was significantly lower than in the group without biliary atresia (median 0.5 [IQR: 0.4–0.7] years vs. 4.9 [IQR: 1.4–12] years, p < 0.001). Biliary complications were described in 52.8% (n = 56) in patients with biliary atresia and in 64.4% (n = 65) of patients with other diagnoses. Significantly more patients with biliary atresia than with other diseases underwent therapy with PTCD (71.4% vs. 40.0%, p < 0.001). ERCP-therapy in biliary atresia-patients was less frequent (32.1% vs. 41.5%, p = 0.347). In total, more biliary atresia-patients received ERCP or PTCD than non-biliary atresia-patients (82% vs. 74%, p = 0.381). Mean BA SDS at the time of LT was 0.89 ± 1.22 in biliary atresia patients vs. -0.50 ± 1.44 in the other group (p = 0.004), at year 1 0.30 ± 1.48 vs. -0.24 ± 1.88 (p = 0.042), at year 3 0.38 ± 1.35 vs. -0.64 ± 1.77 (p < 0.001) and at year 5 0.097 ± 1.084 vs. -0.64 ± 2.11 (p = 0.028). The linear regression model for change in BA SDS from the 3rd year to 5th year showed a significant difference between patients with and without biliary atresia (estimated mean difference β=-0.63, CI [-0.97, -0.29], p < 0.001), adjusting for biliary complications, age and sex, but not for year 1 to year 3 (β = 0.06, CI [-0.34, 0.45], p = 0.773). The change within the first year after LT was not analyzed using a multivariable model due to the large number of missing values. Mean BHI SDS in patients with and without biliary atresia remained negative throughout the observation period. There was a tendency toward higher BHI SDS values in patients with biliary atresia than without (LT: -1.1 ± 0.7 vs. -2.2 ± 1.2, p = 0.001; 1st year: -1.1 ± 0.7 vs. -2.1 ± 1.4 p < 0.001; 3. year: -1.9 ± 1.0 vs. -2.1 ± 1.3, p = 0.407; 5. year: -2.1 ± 1.1 vs. -2.4 ± 1.5, p = 0.255). No statistically significant difference was found between the groups in the multivariable linear regression models for post-LT BHI SDS-change (β=-0.24, CI [-0.67, 0.18], p = 0.256 in 1st-3rd year and β = 0.32, CI [-0.06, 0.69], p = 0.095 in 3rd-5th year). (Fig. 5 ). Biliary complications 121 patients (58.5%) presented with biliary complications post-LT, and 86 (41.5%) without. Of the patients with biliary complications, 66 (54.5%) received treatment with PTCD and 45 (37.2%) with ERCP. Median age of patients with biliary complications was 1.4 (IQR: 0.4–8.6) years vs. 0.7 (IQR: 0.4-5.0) years in patients without (p = 0.33). No sex-related differences were found regarding the presence of biliary complications (p = 0.33) or therapeutic interventions with PTCD or ERCP (p = 0.28). There was no significant difference between the groups in mean BA SDS as well as BHI SDS at all timepoints. Multivariable linear regression analyses in the subgroup of patients with biliary complications and in patients with PTCD/ERCP-therapy vs. without assessing BA SDS and BHI SDS changes post-LT showed no significant difference between the groups. Correlations At all time points, a significant positive correlation was observed between BA SDS and height percentiles (p < 0.001 pre-LT, at year 1 and 3 resp., p = 0.003 at year 5). A positive correlation was estimated between BHI SDS and height percentiles; however, it was only significant pre-LT (p = 0.031). Plasma PTH negatively correlated with BA SDS post-LT. This correlation was rated significant at 3 (p < 0.001) and 5 years (p = 0.003) post-LT. A positive correlation was observed pre-transplant. For BHI SDS, PTH correlated positive prior to LT, negative 1 (not significant, p = 0.38) and 3 (weakly significant, p = 0.047) years post-LT. At year 5, the correlation coefficient was nearly 0 (r = 0.01) (Fig. 6 ). The correlation between BA SDS and BHI SDS was only significant pre-LT (p = 0.034, r = 0.36). A significant inverse association was observed between patient age at LT and the degree of BA SDS reduction post-LT in the linear regression model for the change in BA SDS from year 1 to year 3 (p = 0.002; β=-0.07, CI [-0.12, -0.03]) and year 3 to 5 (p < 0.001; β=-0.13, CI [-0.17, -0.09]), respectively. When controlling for biliary atresia, the effect of age appears to be independent of the underlying condition. For BHI SDS-change post LT, no significant negative correlation was found. All correlation coefficients and p-values are presented in Additional file 1. Fractures The retrospective search for fractures in the clinic’s medical reports revealed 6 patients with fractures. Fractures that could be attributed to underlying osteoporosis were observed in only two patients (Fig. 7 ). Discussion AI applications in pediatric radiology can operate in a fully automated manner, as in the case of BA determination, or in a semi-automated mode. Although BoneXpert, among other methods, was originally developed as AI-replace system, most radiologists are still looking at the hand radiographs to rule out abnormalities 24 . The algorithm is based on the GP method for BA determination 25 , which has been re-evaluated in different populations of the 21st century, indicating no advanced skeletal maturity due to secular change especially in Caucasian/European children 26 , 27 . However, heterogeneities between and within different ethnic groups have been reported 28 , and data show that the GP method should be used with caution in Asian and African populations 29 . To address this limitation, the used software has been and is being validated across Asian, Hispanic, African and Caucasian populations. As there is a complex interrelationship between effects of socioeconomic status and ethnicity on BA estimation, there still remain uncertainties in increasingly heterogeneous populations as the modern central European one. The determination of bone age is important to properly assess and guide the evaluation and, if needed, therapy, of short or tall stature, impaired or accelerated growth, and delayed or early puberty. To the best of our knowledge, this study represents the first systematic evaluation of automated BA and BHI assessment in pLT patients. A significant positive correlation was observed between BA SDS and height percentiles. The tendency to have advanced BA in comparison to chronological age has been reported before in healthy tall children 9 . Patient age at LT was significantly negatively associated with the reduction in BA SDS post-LT, independent of whether the underlying condition was biliary atresia (in constituting a younger age group) or not. The basis for this observation remains to be clarified. Biliary atresia was the single most common indication for pLT in this study, consistent with data from larger study populations 2 , 30 . Pediatric osteoporosis is defined as a clinically significant, low-energy fracture history (clinically significant fracture history is either of the following: ≥2 long-bone fractures by the age of 10 years or ≥ 3 long-bone fractures at any age up to age 19 years) and a bone mineral density (BMD) Z-score ≤ − 2.0, or vertebral compression fractures independent of BMD 31 . The diagnosis of pediatric osteoporosis should not be made based on densitometric criteria (i.e., DXA) alone. Besides ethnicity and genetics, there are complex and multifactorial influences on bone health in children and adolescents, i.e. nutrition and Vitamin D intake, exercise and lifestyle, body weight and composition aswell as hormonal status. In pLT patients, degree of cholestasis and immunosuppressive agents, especially glucocorticoids, can adversely affect bone metabolism 32 . DXA can be useful to detect abnormal BMD and fracture risk in this population but comes with certain limitations in the pediatric population 33 . Furthermore, pediatric-experienced operators and child-specific interpretation are not universally available, and many centers lack pediatric-validated software and reference databases. The assessment of pediatric BMD remains an evolving field, as newer methods are being evaluated in children at risk for poor bone outcomes; among these, quantitative CT (QCT) is not routinely applied in children, but primarily utilized within the context of research studies 4 . At present, no single imaging modality appears capable of providing a comprehensive assessment of bone health in children and adolescents, with each approach exhibiting its own limitations. According to Shalof et al., the digital X-ray radiogrammetry (DXR) method used for BHI assessment achieved the strongest positive relationship with DXA, compared to quantitative ultrasound and peripheral QCT, and should be further evaluated as a predictor of fractures in children and young adults 34 . A significant positive correlation between BHI and BHI-SDS with the DXA and peripheral QCT readings was reported 35 . DXR is observer-independent, inexpensive, widely available and offers the advantage of a very low radiation dose of < 0.00012 mSv per hand radiograph 8 . Automated adult height prediction based on BA assessment can be combined, contributing to its effectiveness in clinical practice. Reduced bone density is a common complication of chronic liver disease in both adults and children and relates to the presence of cholestasis 36 . In this study, mean BHI SDS was reduced at all timepoints and notably, declined progressively until year 5 post-LT, independent of the underlying disease. This could indicate a reduced and further decreasing BMD in this cohort of pLT recipients. The quite proactive treatment approach of biliary complications including cholestatic disease post-LT could provide an explanation for why no difference exists between the groups with and without biliary complications. It should be noted that cholestasis cannot be assessed solely by radiological means and does not consistently manifest as visible dilatation of the bile ducts. In pathologic states like chronic kidney disease, higher PTH tends to correlate with poorer bone mineralisation 37 . A previous study involving children with chronic kidney disease showed reduced BHI and a weakly significant negative correlation between BHI SDS and plasma PTH 38 . Within the present study, the correlation between BHI SDS and plasma PTH was found to be negative at year 1 (not significant) and 3 (weakly significant) post-LT; also, there was a negative correlation between BA SDS and plasma PTH post-LT. Notably, the relationship between PTH and bone density in pediatric chronic liver disease is not as straightforward as seen in adult hyperparathyroidism. Several pediatric studies do not find significantly elevated PTH in liver disease; deficits in BMD in these children seem to be more complex and multifactorial, and PTH cannot be considered as main correlate of low BMD in pediatric liver disease or pLT recipients. According to previous data, the overall risk of sustaining a fracture during childhood is 10–25% and the lifetime risk of sustaining a fracture 27% for girls and 42% for boys 39 . The relationship between fracture risk and BHI has been described in specific disease groups (like Duchenne muscle dystrophy). The small number of fractures in our cohort is likely underestimated within the framework of the retrospective data analysis. In this center, no screening for fractures had been carried out, as described elsewhere 5 . In a recent study, an epidemiological association of BHI with both BMD and fracture risk was described, and the authors suggested that automated BHI assessment using hand radiographs might constitute a cost-effective measurement to identify children at risk of low BMD and fractures in pediatric clinics 40 . This study has some limitations due to its retrospective and monocentric design. The study population is partially heterogeneous, including some rare underlying diseases. In addition, the small number of preoperative radiographs compared to postoperative ones considerably limited the comparison between these two time points. Although bone age and final height assessments are routinely performed in pLT recipients, a potential bias might exist due to the selection of patients who underwent radiographic examination. As DXA is not routinely performed in these patients, a comparison between the methods is not possible. Further studies with a longer follow-up period will be required to characterize bone health and assess fracture risk in pLT recipients. Conclusions Whereas mean BA SDS remained approximately normotypic following pLT, mean BHI SDS was persistently reduced and exhibited a continuous decline through year 5. This may suggest a progressive decline in BMD post-LT. Further prospective studies with larger cohorts are needed to determine the utility of fully automated BA and BHI assessment in pLT recipients. Abbreviations AI: artificial intelligence AP: anterio-posterior BA SDS: bone age standard deviation score BHI SDS: bone health index standard deviation score BMD: bone mineral density CI: confidence interval DXA: dual energy X-ray absorptiometry DV: dorso-volar GP: Greulich and Pyle pLT: pediatric liver transplantation PTCD: percutaneous transhepatic cholangiographic drainage ERCP: endoscopic retrograde cholangiopancreatography PTH: parathyroid hormone QCT: quantitative CT Declarations Statement of contribution BG contributed to the research design, writing the paper, performance of the research and data analysis. All authors had full access to the data. GN contributed to data acquisition and writing the paper. FK contributed to data analysis. SH, QS, KB, CS contributed to revising the draft. GN contributed to the conception and design of the work and data analysis. SH contributed to writing the paper and final approval of the paper. Human Ethics and Consent to Participate declarations This study was performed in accordance with the ethical principles of the Declaration of Helsinki and its amendments and with the approval of the local IRB („Ethics Comittee, University Regensburg“; vote number: 24-4012-104). Written informed consent was waived by the Institutional Review Board owing to the retrospective nature of the evaluation of radiographs acquired in routine clinical practice. Consent for Publication Not applicable. Declaration of Competing Interest All authors declare no conflicts of interest. Funding source The authors received no financial support for the research, authorship and publication of this article. The scientific department of Bonexpert provided the software for retrospective analysis free of charge. Data sharing statement The data that support the findings of this study are available on request from the corresponding author. References Vimalesvaran S, Verma A, Dhawan A. Pediatric Liver Transplantation: Selection Criteria and Post-transplant Medical Management. Indian J Pediatr. 2024;91(4):383–90. 10.1007/s12098-023-04963-5 . Baumann U, Karam V, Adam R, et al. Prognosis of Children Undergoing Liver Transplantation: A 30-Year European Study. Pediatrics. 2022;150(4). 10.1542/peds.2022-057424 . Vimalesvaran S, Souza LN, Deheragoda M, et al. Outcomes of adults who received liver transplant as young children. EClinicalMedicine. 2021;38:100987. 10.1016/j.eclinm.2021.100987 . Kusumi K, Shaikhkhalil A, Patel HP, Mahan JD. Promoting bone health in children and adolescents following solid organ transplantation. Pediatr Transpl. 2021;25(1):e13940. 10.1111/petr.13940 . Helenius I, Remes V, Salminen S, et al. Incidence and predictors of fractures in children after solid organ transplantation: a 5-year prospective, population-based study. J Bone Min Res. 2006;21(3):380–7. 10.1359/JBMR.051107 . Anastasilakis AD, Tsourdi E, Makras P, et al. Bone disease following solid organ transplantation: A narrative review and recommendations for management from The European Calcified Tissue Society. Bone. 2019;127:401–18. 10.1016/j.bone.2019.07.006 . Lim WH, Ng CH, Ow ZGW, et al. A systematic review and meta-analysis on the incidence of osteoporosis and fractures after liver transplant. Transpl Int. 2021;34(6):1032–43. 10.1111/tri.13863 . Cavallo F, Mohn A, Chiarelli F, Giannini C. Evaluation of Bone Age in Children: A Mini-Review. Front Pediatr. 2021;9:580314. 10.3389/fped.2021.580314 . Martin DD, Wit JM, Hochberg Z, et al. The use of bone age in clinical practice - part 1. Horm Res Paediatr. 2011;76(1):1–9. 10.1159/000329372 . Martin DD, Meister K, Schweizer R, Ranke MB, Thodberg HH, Binder G. Validation of automatic bone age rating in children with precocious and early puberty. Offiah AC. Current and emerging artificial intelligence applications for pediatric musculoskeletal radiology. Pediatr Radiol. 2022;52(11):2149–58. 10.1007/s00247-021-05130-8 . Martin DD, Calder AD, Ranke MB, Binder G, Thodberg HH. Accuracy and self-validation of automated bone age determination. Sci Rep. 2022;12(1):6388. 10.1038/s41598-022-10292-y . Martin DD, Heckmann C, Jenni OG, Ranke MB, Binder G, Thodberg HH. Metacarpal thickness, width, length and medullary diameter in children—reference curves from the First Zürich Longitudinal Study. Osteoporos Int. 2011;22(5):1525–36. 10.1007/s00198-010-1389-9 . Bonnard GD. Cortical thickness and diaphysial diameter of the metacarpal bones from the age of three months to eleven years. Helv Paediatr Acta. 1968;23(5):445–63. Breen MA, Tsai A, Stamm A, Kleinman PK. Bone age assessment practices in infants and older children among Society for Pediatric Radiology members. Pediatr Radiol. 2016;46(9):1269–74. 10.1007/s00247-016-3618-7 . https://bonexpert.com. Accessed 13 January2026. Thodberg HH, Kreiborg S, Juul A, Pedersen KD. The BoneXpert method for automated determination of skeletal maturity. IEEE Trans Med Imaging. 2009;28(1):52–66. 10.1109/TMI.2008.926067 . Thodberg HH, van Rijn RR, Jenni OG, Martin DD. Automated determination of bone age from hand X-rays at the end of puberty and its applicability for age estimation. Int J Legal Med. 2017;131(3):771–80. 10.1007/s00414-016-1471-8 . Thodberg HH, van Rijn RR, Tanaka T, Martin DD, Kreiborg S. A paediatric bone index derived by automated radiogrammetry. Osteoporos Int. 2010;21(8):1391–400. 10.1007/s00198-009-1085-9 . Neubert K, Brunner E. A studentized permutation test for the non-parametric Behrens–Fisher problem. Comput Stat Data Anal. 2007;51(10):5192–204. 10.1016/j.csda.2006.05.024 . Konietschke F, Placzek M, Schaarschmidt F, Hothorn LA. nparcomp: An R Software Package for Nonparametric Multiple Comparisons and Simultaneous Confidence Intervals. J Stat Soft. 2015;64(9):1–17. 10.18637/jss.v064.i09 . Patil I. Visualizations with statistical details: The 'ggstatsplot' approach. JOSS. 2021;6(61):3167. 10.21105/joss.03167 . Lüdecke D. ggeffects: Tidy Data Frames of Marginal Effects from Regression Models. JOSS. 2018;3(26):772. 10.21105/joss.00772 . Thodberg HH, Thodberg B, Ahlkvist J, Offiah AC. Autonomous artificial intelligence in pediatric radiology: the use and perception of BoneXpert for bone age assessment. Pediatr Radiol. 2022;52(7):1338–46. 10.1007/s00247-022-05295-w . GREULICH WW, PYLE SI. RADIOGRAPHIC ATLAS OF SKELETAL DEVELOPMENT OF, THE HAND AND WRIST. Am J Med Sci. 1959;238(3):393. 10.1097/00000441-195909000-00030 . Alshamrani K, Offiah AC. Applicability of two commonly used bone age assessment methods to twenty-first century UK children. Eur Radiol. 2020;30(1):504–13. 10.1007/s00330-019-06300-x . Dahlberg PS, Mosdøl A, Ding Y, et al. A systematic review of the agreement between chronological age and skeletal age based on the Greulich and Pyle atlas. Eur Radiol. 2019;29(6):2936–48. 10.1007/s00330-018-5718-2 . Mansourvar M, Ismail MA, Raj RG, et al. The applicability of Greulich and Pyle atlas to assess skeletal age for four ethnic groups. J Forensic Leg Med. 2014;22:26–9. 10.1016/j.jflm.2013.11.011 . Alshamrani K, Messina F, Offiah AC. Is the Greulich and Pyle atlas applicable to all ethnicities? A systematic review and meta-analysis. Eur Radiol. 2019;29(6):2910–23. 10.1007/s00330-018-5792-5 . Junge N, Karam V, Hartog H, et al. Update on pediatric liver transplantation in Europe 2022: An ELITA-ESPGHAN report. J Pediatr Gastroenterol Nutr. 2025;81(1):82–90. 10.1002/jpn3.70065 . Weber DR, Boyce A, Gordon C, et al. The Utility of DXA Assessment at the Forearm, Proximal Femur, and Lateral Distal Femur, and Vertebral Fracture Assessment in the Pediatric Population: 2019 ISCD Official Position. J Clin Densitom. 2019;22(4):567–89. 10.1016/j.jocd.2019.07.002 . Jang MJ, Shin C, Kim S, et al. Factors affecting bone mineral density in children and adolescents with secondary osteoporosis. Ann Pediatr Endocrinol Metab. 2023;28(1):34–41. 10.6065/apem.2244026.013 . Khalatbari H, Binkovitz LA, Parisi MT. Dual-energy X-ray absorptiometry bone densitometry in pediatrics: a practical review and update. Pediatr Radiol. 2021;51(1):25–39. 10.1007/s00247-020-04756-4 . Shalof H, Dimitri P, Shuweihdi F, Offiah AC. Which skeletal imaging modality is best for assessing bone health in children and young adults compared to DXA? A systematic review and meta-analysis. Bone. 2021;150:116013. 10.1016/j.bone.2021.116013 . Schündeln MM, Marschke L, Bauer JJ, et al. A Piece of the Puzzle: The Bone Health Index of the BoneXpert Software Reflects Cortical Bone Mineral Density in Pediatric and Adolescent Patients. PLoS ONE. 2016;11(3):e0151936. 10.1371/journal.pone.0151936 . Loomes KM, Spino C, Goodrich NP, et al. Bone Density in Children With Chronic Liver Disease Correlates With Growth and Cholestasis. Hepatology. 2019;69(1):245–57. 10.1002/hep.30196 . Lalayiannis AD, Crabtree NJ, Fewtrell M, et al. Assessing bone mineralisation in children with chronic kidney disease: what clinical and research tools are available? Pediatr Nephrol. 2020;35(6):937–57. 10.1007/s00467-019-04271-1 . Nüsken E, Imschinetzki D, Nüsken K-D, et al. Automated Greulich-Pyle bone age determination in children with chronic kidney disease. Pediatr Nephrol. 2015;30(7):1173–9. 10.1007/s00467-015-3042-5 . Larsen AV, Mundbjerg E, Lauritsen JM, Faergemann C. Development of the annual incidence rate of fracture in children 1980–2018: a population-based study of 32,375 fractures. Acta Orthop. 2020;91(5):593–7. 10.1080/17453674.2020.1772555 . Prijatelj V, Grgic O, Uitterlinden AG, Wolvius EB, Rivadeneira F, Medina-Gomez C. Bone health index in the assessment of bone health: The Generation R Study. Bone. 2024;182:117070. 10.1016/j.bone.2024.117070 . Additional Declarations No competing interests reported. Supplementary Files AdditionalFile1.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 06 Mar, 2026 Editor assigned by journal 04 Mar, 2026 Editor invited by journal 04 Feb, 2026 Submission checks completed at journal 04 Feb, 2026 First submitted to journal 04 Feb, 2026 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-8694835","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":602782307,"identity":"3eb5183e-61b1-44fd-8f95-5a89513c8c8d","order_by":0,"name":"Gerardo Napodano","email":"","orcid":"","institution":"University Hospital Regensburg","correspondingAuthor":false,"prefix":"","firstName":"Gerardo","middleName":"","lastName":"Napodano","suffix":""},{"id":602782308,"identity":"73d80791-1cfe-44af-8946-41d142c7038e","order_by":1,"name":"Fabian Kück","email":"","orcid":"","institution":"University of Göttingen","correspondingAuthor":false,"prefix":"","firstName":"Fabian","middleName":"","lastName":"Kück","suffix":""},{"id":602782309,"identity":"ffa1fe02-41bc-4811-bd3d-cbdb13f4d5ef","order_by":2,"name":"Quirin Strotzer","email":"","orcid":"","institution":"University Hospital Regensburg","correspondingAuthor":false,"prefix":"","firstName":"Quirin","middleName":"","lastName":"Strotzer","suffix":""},{"id":602782310,"identity":"12258ab1-6753-4a35-8b99-0324fbcb4e9c","order_by":3,"name":"Birgit Knoppke","email":"","orcid":"","institution":"University Hospital Regensburg","correspondingAuthor":false,"prefix":"","firstName":"Birgit","middleName":"","lastName":"Knoppke","suffix":""},{"id":602782311,"identity":"d7ec5d8a-2106-44cb-87f6-e46fa7544374","order_by":4,"name":"Christian Stroszczynski","email":"","orcid":"","institution":"University Hospital Regensburg","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Stroszczynski","suffix":""},{"id":602782312,"identity":"02c72c5e-4d76-4569-9a4b-2e68240f7ffd","order_by":5,"name":"Simone Hammer","email":"","orcid":"","institution":"University Hospital Regensburg","correspondingAuthor":false,"prefix":"","firstName":"Simone","middleName":"","lastName":"Hammer","suffix":""},{"id":602782313,"identity":"3fe6df7b-311a-4af2-a181-c479acfcb58d","order_by":6,"name":"Barbara Greiner","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/klEQVRIiWNgGAWjYBACAwYGNjjnABDLMfAASR5StBiTpgUEEhsIaTFnP/vswc89dgzy0YcPHi6ouJe+4czZBwxvKnBrsexJNzfseZbMYHguLeHwjDPFuRvOthswzjmDx2EH0tgkeA4wMxj28Bgc5m1LyN1wno2BmbcNj5bzz9gk/xyoB2rh/3CY919CugFYyz88Wm6ksUnzHDjMIM/Dw3CYtyEhweBsG1BLAx6/zHjGbixz4DiPAQ+bwWGeYwmGM88cYzg45xhuLeb8aWwP3xyolpPvYX78macmQZ7vTBrjgzc1uLXAAI/BASTeARyqUIE8HtePglEwCkbBCAcABwFPEL7xXqUAAAAASUVORK5CYII=","orcid":"","institution":"University Hospital Regensburg","correspondingAuthor":true,"prefix":"","firstName":"Barbara","middleName":"","lastName":"Greiner","suffix":""}],"badges":[],"createdAt":"2026-01-25 20:53:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8694835/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8694835/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104413854,"identity":"a0cc1172-90a1-4eb1-b2da-806d369628f4","added_by":"auto","created_at":"2026-03-11 13:05:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":115165,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of patient selection, detailing inclusion and exclusion criteria and corresponding patient numbers.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/88390bde2e523d6d0147432e.png"},{"id":104411945,"identity":"f0971df6-fef5-4354-aa76-dd2ccdf8275c","added_by":"auto","created_at":"2026-03-11 12:58:11","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":336735,"visible":true,"origin":"","legend":"\u003cp\u003eExemplary output of automated BA assessment in patient with atypical biliary atresia one year post-LT, PTCD-therapy due to cholestasis and biloma. BA (GP): Bone Age (Greulich\u0026amp;Pyle). BA (TW): BA (Tanner-Whitehouse) BA SDS (Standard deviation score, also known as Z-score). BHI: Bone Health Index. Ethnicity: Caucasian/European. No age specified for data protection reasons.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/b93c6a3f735d83505e2e1c20.png"},{"id":104411947,"identity":"799e0d8f-7b33-46b0-8553-be4b5a8a898a","added_by":"auto","created_at":"2026-03-11 12:58:11","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":25708,"visible":true,"origin":"","legend":"\u003cp\u003eDetermination of the inner and outer metacarpal cortex and calculation of BHI. L= total metacarpal length.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/a30cc1821004acbd64e956e1.png"},{"id":104411996,"identity":"f11551cd-560e-4ec8-a987-9afeabc026f4","added_by":"auto","created_at":"2026-03-11 12:58:26","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":397640,"visible":true,"origin":"","legend":"\u003cp\u003evisualization of temporal progression of BA SDS (a) and BHI SDS (b).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/d40e244611cc43526db821f2.png"},{"id":104414048,"identity":"90ddeb53-ae7f-4aa0-a107-bc0b9d4864f1","added_by":"auto","created_at":"2026-03-11 13:06:25","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":427902,"visible":true,"origin":"","legend":"\u003cp\u003eMultivariable linear regression models for post-LT BA SDS and BHI SDS-change in patients with and without biliary atresia.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/8bd6673a8636aba30c6572bd.png"},{"id":104415559,"identity":"83a03523-07a8-45ad-891c-859be6dff110","added_by":"auto","created_at":"2026-03-11 13:11:13","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":422609,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelations between BA SDS and BHI SDS and height precentiles as well as plasma PTH.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/f669c672d1c11e8d9725e769.png"},{"id":104412007,"identity":"03738880-a3ed-4e1e-b112-036b350d40bc","added_by":"auto","created_at":"2026-03-11 12:58:28","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":235989,"visible":true,"origin":"","legend":"\u003cp\u003eToddler with Alagille`s syndrome (caused by NOTCH2 mutation, known for associated osteopathy) and severe malnutrition post-LT. Multiple spontaneous vertebral and left femoral fractures.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/c86f5fc2bfbdab29bd7463e6.png"},{"id":104779958,"identity":"e6e7e50e-d20b-4399-ad7d-f61ec6a34f10","added_by":"auto","created_at":"2026-03-17 07:48:30","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2481029,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/c022c294-176a-4e4f-9617-8fa217358c45.pdf"},{"id":104414144,"identity":"14fbdbba-d612-4d63-a350-5ea4a70bce88","added_by":"auto","created_at":"2026-03-11 13:06:53","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":17019,"visible":true,"origin":"","legend":"","description":"","filename":"AdditionalFile1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8694835/v1/5d316d7335d3792cc2c36582.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Automated determination of bone age and bone health index in pediatric liver transplant recipients","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePediatric liver transplantation (pLT) is the gold standard for life-threatening acute and chronic liver diseases and multiple liver-based inherited metabolic defects and should be considered for all children with end-stage liver disease \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Since long-term patient and graft survival are excellent, attainment of normal growth and development are crucial to support children in leading a life that is as close to normal as possible and to be an independent adult. This is particularly relevant for very young pLT recipients, where infantile-onset liver disease and therapeutic measures, in particular surgery and immunosuppression, can adversely affect growth and neurodevelopment \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePediatric bone status entering adulthood is considered a critical determinant of bone health through adulthood. Bone status may be adversely affected due to pre-existing bone disease, effects of medications and nutritional challenges related to pLT \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Elevated fracture risk in patients following liver and other solid organ transplantation has been described in children as well as adults \u003csup\u003e\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Determination of bone age (BA) to assess skeletal maturity and predict adult height is applied in various pediatric diseases and their associated therapeutic interventions, including chronic liver diseases \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Since manual BA determination is time-consuming and associated with considerable inter-rater variability \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, fully automated image analysis methods have been developed. The method uses a validated machine-learning approach for BA prediction. It has been shown to outperform single manual readings and to produce fewer major errors, indicating that it may be safer than manual assessment\u003csup\u003e11 12\u003c/sup\u003e. Another promising biomarker of skeletal health in children is the Bone Health Index (BHI) that is based on the automated assessment of metacarpal thickness, width, length and medullary diameter \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAt our institution, the first pLTs were performed in 2008 and have been ongoing ever since. Normal patient follow-up care takes place at the outpatient LT clinic and data obtained here is recorded in the pLT register. Left hand radiographs have been taken during annual outpatient visits to determine BA and its progression. Given the absence of systematic data on automated BA and BHI assessments in pLT patients, our study examined whether BA SDS and BHI SDS deviate from normative standards in this population.\u003c/p\u003e"},{"header":"Material and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy design and patient population\u003c/h2\u003e \u003cp\u003e This retrospective single-center study was performed in accordance with the ethical principles of the Declaration of Helsinki and its amendments and with the approval of the local IRB (vote number: 24-4012-104). The IRB waived the requirement for written consent.\u003c/p\u003e \u003cp\u003eThe local picture archiving and communication system (PACS) was queried for hand radiographs (a.p. or d.v.) of patients\u0026thinsp;\u0026lt;\u0026thinsp;18 years, acquired between January 1, 2008 and April 30, 2024, and whose reports included the search terms \u0026lsquo;liver transplant\u0026rsquo;, \u0026lsquo;LT\u0026rsquo;, or \u0026lsquo;bone age\u0026rsquo;. This yielded 324 patients, of whom 31 were excluded because no liver transplant had been performed or no post-transplant radiograph was available.\u003c/p\u003e \u003cp\u003eThe data for the remaining 293 individuals were compared with the entries in the local pLT-registry that is based on the password-encrypted QNOME study platform. 50 individuals were not registered due to various reasons (i.e., pLT in patients who had turned 18, loss to follow-up) and were therefore excluded. 36 patients were ineligible due to unsuitable timing of the X-ray examination. (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFor the remaining 207 patients who met the inclusion criteria, descriptive data were extracted from the registry, local PACS and hospital information system. Data collection of pre-transplant demographics included underlying disease, age and sex of recipient at LT; post-transplant details explored included biliary complications, PTCD and/or ERCP-therapy (including bile duct stenting), and fractures. Various types of complications are subsumed in the register under the umbrella term \u0026bdquo;biliary complications\u0026ldquo;: intrahepatic duct stenosis (at the anastomosis or extrahepatic); biliary leckage/bilioma; cholangitis; ischemic type biliary lesion; biliary cast/sludge; secondary sclerosing cholangitis; necrosis; others.\u003c/p\u003e \u003cp\u003eHeight percentile and plasma PTH at different timepoints (preparation for surgery and 1-, 3-, and 5-years post LT) were noted.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAutomated bone age estimation\u003c/h3\u003e\n\u003cp\u003eBoneXpert has been used for automated, PACS-integrated bone age assessment at our institution since May 2023. All images relevant to the study that were taken before this date were retrospectively evaluated using a standalone solution tailored for research purposes. BoneXpert automatically calculates bone age according to the Greulich and Pyle and Tanner Whitehouse standards in a process that takes less than 15 s per radiograph via a Digital Imaging and Communications in Medicine (DICOM) node. No data are stored, shared or transferred outside the local PACS.\u003c/p\u003e \u003cp\u003eChronological age, BA according to Greulich and Pyle (GP), the most common used bone age estimation technique \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, and BA SDS, BHI and BHI SDS were documented for each radiograph (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eBone Age. BA SDS\u0026thinsp;\u0026gt;\u0026thinsp;0 indicates advanced bone age, BA SDS\u0026thinsp;\u0026lt;\u0026thinsp;0 indicates delayed bone age. The overall reported BA (GP) accuracy is 0.62 years, with a root mean square error and a mean absolute deviation of 0.30 and 0.21 years in infants, respectively\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. The underlying mathematical framework has been described previously \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. An extension of the method at the end of puberty, up to a BA of 19 years for boys and 18 years for girls of Caucasian (most prevalent ethnicity in this study cohort) origin has been reported \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBone Health Index. Accuracy is defined as the degree of agreement with a reference method. Since BHI represents a concept derived by automated radiogrammetry, no such external reference exists. Consequently, BHI values are expressed in arbitrary units. Interpretation relies on comparison with a reference curve, which serves as the internal reference standard. As this curve is obtained using the same device, BHI accuracy relative to the reference equals the method\u0026rsquo;s precision, characterized by a relative SD of 1.4% at a cortical thickness of 1.3 mm (appr. 10 years of age) \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The underlying parameters and methodological analysis have been discussed elsewhere \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eVariables were summarized using absolute and relative frequencies, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation, median, range and interquartile range (IQR).\u003c/p\u003e \u003cp\u003eFor the comparison of different groups, we applied Fisher\u0026rsquo;s exact test for nominal variables, Welch\u0026rsquo;s t test for BA SDS and BHI SDS, and the non-parametric Studentized permutation test by Neubert and Brunner for all other numeric variables\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Moreover, we computed Pearson correlation coefficients and fitted multivariable linear regression models. Results are reported with 95% confidence intervals (CI) and visualized using estimated marginal means.\u003c/p\u003e \u003cp\u003eThe significance level was set to α= 5% for all statistical tests. Due to the exploratory nature of this study, no adjustment for multiple testing was applied. All analyses were performed with the statistical programming environment R (version 4.2.3; R Core Team (2023)) using the R packages nparcomp (version 3.0) \u003csup\u003e21\u003c/sup\u003e for the studentized permutation test, and ggstatsplot (version 0.12.0)\u003csup\u003e22\u003c/sup\u003e and ggeffects (version 2.3.1) \u003csup\u003e23\u003c/sup\u003e for the visualizations.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e106 patients (51.2%) presented with biliary atresia, 101 (48.8%) with other diagnoses, the most common of which being non A-E hepatitis (6.8%), progressive familial intrahepatic cholestasis (4.8%), biliary hypoplasia in Alagille`s disease (3.9%) and cystic fibrosis (3.4%).\u003c/p\u003e \u003cp\u003e104 patients (50.2%) were female, 103 male (49.8%). Median age at LT was 1.1 years (IQR: 0.43; 6.8 years). Mean age at the time of LT was 4.1\u0026thinsp;\u0026plusmn;\u0026thinsp;5.2 years. The standard deviation is attributable to the fact that in this group, many children are very young, but some significantly older children are included, explaining the high dispersion. 83% of the recipients had received a split liver graft, 17% a whole organ. 23 patients died between 2008 and 2024. Standard immunosuppressive therapy included corticosteroid administration. All pLT patients underwent serum level\u0026ndash;guided vitamin D supplementation. Determination of BA and its progression was advised in patients pre-LT except for neonates, and post-LT at the annual outpatient visits except for patients in whom epiphyseal closure had already occurred.\u003c/p\u003e \u003cp\u003e449 radiographs were evaluated. 13 images were automatically rejected. For the rejected images, BHI could not be determined, and BA was evaluated manually. There were 37 images prior to LT, 167 at year 1, 128 at year 3 and 117 at year 5.\u003c/p\u003e\n\u003ch3\u003eBA SDS and BHI SDS\u003c/h3\u003e\n\u003cp\u003eMean BA SDS remained around the zero value at all time points in the overall cohort (LT: 0.03\u0026thinsp;\u0026plusmn;\u0026thinsp;1.50; 1 year: 0.03\u0026thinsp;\u0026plusmn;\u0026thinsp;1.70; 3 years: \u0026minus;0.07\u0026thinsp;\u0026plusmn;\u0026thinsp;1.62; 5 years: \u0026minus;0.21\u0026thinsp;\u0026plusmn;\u0026thinsp;1.63; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). The difference from LT to year 1 was significant (mean\u0026thinsp;\u0026minus;\u0026thinsp;0.57\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22, p\u0026thinsp;=\u0026thinsp;0.022) in the subset with available paired measurements; however, patients with an available BA measurement at LT had lower BA SDS at one year than those without a baseline measurement. As with BA, patients with an available BHI measurement at the time of LT had lower BHI at 1 year than those without a baseline measurement. The difference between year 1 and 3 (mean\u0026thinsp;\u0026minus;\u0026thinsp;0.01\u0026thinsp;\u0026plusmn;\u0026thinsp;0.97, p\u0026thinsp;=\u0026thinsp;0.310) and between 3 and 5 (mean\u0026thinsp;\u0026minus;\u0026thinsp;0.13\u0026thinsp;\u0026plusmn;\u0026thinsp;0.80, p\u0026thinsp;=\u0026thinsp;0.117) was not significant.\u003c/p\u003e \u003cp\u003eMean BHI SDS at the time of LT was \u0026minus;\u0026thinsp;1.8\u0026thinsp;\u0026plusmn;\u0026thinsp;1.20, -1.6\u0026thinsp;\u0026plusmn;\u0026thinsp;1.20 one year post LT, -2.0\u0026thinsp;\u0026plusmn;\u0026thinsp;1.10 at the third-year check-up and \u0026minus;\u0026thinsp;2.2\u0026thinsp;\u0026plusmn;\u0026thinsp;1.3 five years post LT (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). The change in BHI SDS between the 1st and 3rd year (mean\u0026thinsp;\u0026minus;\u0026thinsp;0.40\u0026thinsp;\u0026plusmn;\u0026thinsp;1.10) as well as between 3rd and 5th year (mean\u0026thinsp;\u0026minus;\u0026thinsp;0.16\u0026thinsp;\u0026plusmn;\u0026thinsp;0.71) was significant with a p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.001 and 0.032, respectively. As with BA, patients with a BHI measurement at LT had lower mean BHI SDS at 1 year than those without a baseline measurement; however, the difference from LT to year 1 (mean\u0026thinsp;\u0026minus;\u0026thinsp;0.19\u0026thinsp;\u0026plusmn;\u0026thinsp;0.73) was not statistically significant in the subset with available paired measurements (p\u0026thinsp;=\u0026thinsp;0.204) with a relatively wide CI reflecting the limited subgroup size (n\u0026thinsp;=\u0026thinsp;26, 95% CI [-0.48; 0.11]). BA SDS and BHI SDS were not found to be significantly influenced by patients' sex.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eBiliary atresia\u003c/h2\u003e \u003cp\u003eIn the 106 patients with biliary atresia, the proportion of females (54.7%) was greater than that of males (45.3%) compared to patients with other underlying diseases (45.5% female and 54.5% male) with no statistically significant difference (p\u0026thinsp;=\u0026thinsp;0.212). Mean age was significantly lower than in the group without biliary atresia (median 0.5 [IQR: 0.4\u0026ndash;0.7] years vs. 4.9 [IQR: 1.4\u0026ndash;12] years, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eBiliary complications were described in 52.8% (n\u0026thinsp;=\u0026thinsp;56) in patients with biliary atresia and in 64.4% (n\u0026thinsp;=\u0026thinsp;65) of patients with other diagnoses. Significantly more patients with biliary atresia than with other diseases underwent therapy with PTCD (71.4% vs. 40.0%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). ERCP-therapy in biliary atresia-patients was less frequent (32.1% vs. 41.5%, p\u0026thinsp;=\u0026thinsp;0.347). In total, more biliary atresia-patients received ERCP or PTCD than non-biliary atresia-patients (82% vs. 74%, p\u0026thinsp;=\u0026thinsp;0.381).\u003c/p\u003e \u003cp\u003eMean BA SDS at the time of LT was 0.89\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22 in biliary atresia patients vs. -0.50\u0026thinsp;\u0026plusmn;\u0026thinsp;1.44 in the other group (p\u0026thinsp;=\u0026thinsp;0.004), at year 1 0.30\u0026thinsp;\u0026plusmn;\u0026thinsp;1.48 vs. -0.24\u0026thinsp;\u0026plusmn;\u0026thinsp;1.88 (p\u0026thinsp;=\u0026thinsp;0.042), at year 3 0.38\u0026thinsp;\u0026plusmn;\u0026thinsp;1.35 vs. -0.64\u0026thinsp;\u0026plusmn;\u0026thinsp;1.77 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and at year 5 0.097\u0026thinsp;\u0026plusmn;\u0026thinsp;1.084 vs. -0.64\u0026thinsp;\u0026plusmn;\u0026thinsp;2.11 (p\u0026thinsp;=\u0026thinsp;0.028). The linear regression model for change in BA SDS from the 3rd year to 5th year showed a significant difference between patients with and without biliary atresia (estimated mean difference β=-0.63, CI [-0.97, -0.29], p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), adjusting for biliary complications, age and sex, but not for year 1 to year 3 (β\u0026thinsp;=\u0026thinsp;0.06, CI [-0.34, 0.45], p\u0026thinsp;=\u0026thinsp;0.773). The change within the first year after LT was not analyzed using a multivariable model due to the large number of missing values.\u003c/p\u003e \u003cp\u003eMean BHI SDS in patients with and without biliary atresia remained negative throughout the observation period. There was a tendency toward higher BHI SDS values in patients with biliary atresia than without (LT: -1.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.7 vs. -2.2\u0026thinsp;\u0026plusmn;\u0026thinsp;1.2, p\u0026thinsp;=\u0026thinsp;0.001; 1st year: -1.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.7 vs. -2.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.4 p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; 3. year: -1.9\u0026thinsp;\u0026plusmn;\u0026thinsp;1.0 vs. -2.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.3, p\u0026thinsp;=\u0026thinsp;0.407; 5. year: -2.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1 vs. -2.4\u0026thinsp;\u0026plusmn;\u0026thinsp;1.5, p\u0026thinsp;=\u0026thinsp;0.255). No statistically significant difference was found between the groups in the multivariable linear regression models for post-LT BHI SDS-change (β=-0.24, CI [-0.67, 0.18], p\u0026thinsp;=\u0026thinsp;0.256 in 1st-3rd year and β\u0026thinsp;=\u0026thinsp;0.32, CI [-0.06, 0.69], p\u0026thinsp;=\u0026thinsp;0.095 in 3rd-5th year). (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eBiliary complications\u003c/h3\u003e\n\u003cp\u003e121 patients (58.5%) presented with biliary complications post-LT, and 86 (41.5%) without. Of the patients with biliary complications, 66 (54.5%) received treatment with PTCD and 45 (37.2%) with ERCP. Median age of patients with biliary complications was 1.4 (IQR: 0.4\u0026ndash;8.6) years vs. 0.7 (IQR: 0.4-5.0) years in patients without (p\u0026thinsp;=\u0026thinsp;0.33). No sex-related differences were found regarding the presence of biliary complications (p\u0026thinsp;=\u0026thinsp;0.33) or therapeutic interventions with PTCD or ERCP (p\u0026thinsp;=\u0026thinsp;0.28).\u003c/p\u003e \u003cp\u003eThere was no significant difference between the groups in mean BA SDS as well as BHI SDS at all timepoints. Multivariable linear regression analyses in the subgroup of patients with biliary complications and in patients with PTCD/ERCP-therapy vs. without assessing BA SDS and BHI SDS changes post-LT showed no significant difference between the groups.\u003c/p\u003e"},{"header":"Correlations","content":"\u003cp\u003eAt all time points, a significant positive correlation was observed between BA SDS and height percentiles (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001 pre-LT, at year 1 and 3 resp., p\u0026thinsp;=\u0026thinsp;0.003 at year 5). A positive correlation was estimated between BHI SDS and height percentiles; however, it was only significant pre-LT (p\u0026thinsp;=\u0026thinsp;0.031). Plasma PTH negatively correlated with BA SDS post-LT. This correlation was rated significant at 3 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and 5 years (p\u0026thinsp;=\u0026thinsp;0.003) post-LT. A positive correlation was observed pre-transplant. For BHI SDS, PTH correlated positive prior to LT, negative 1 (not significant, p\u0026thinsp;=\u0026thinsp;0.38) and 3 (weakly significant, p\u0026thinsp;=\u0026thinsp;0.047) years post-LT. At year 5, the correlation coefficient was nearly 0 (r\u0026thinsp;=\u0026thinsp;0.01) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe correlation between BA SDS and BHI SDS was only significant pre-LT (p\u0026thinsp;=\u0026thinsp;0.034, r\u0026thinsp;=\u0026thinsp;0.36).\u003c/p\u003e \u003cp\u003eA significant inverse association was observed between patient age at LT and the degree of BA SDS reduction post-LT in the linear regression model for the change in BA SDS from year 1 to year 3 (p\u0026thinsp;=\u0026thinsp;0.002; β=-0.07, CI [-0.12, -0.03]) and year 3 to 5 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; β=-0.13, CI [-0.17, -0.09]), respectively. When controlling for biliary atresia, the effect of age appears to be independent of the underlying condition. For BHI SDS-change post LT, no significant negative correlation was found. All correlation coefficients and p-values are presented in Additional file 1.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eFractures\u003c/h2\u003e \u003cp\u003eThe retrospective search for fractures in the clinic\u0026rsquo;s medical reports revealed 6 patients with fractures. Fractures that could be attributed to underlying osteoporosis were observed in only two patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eAI applications in pediatric radiology can operate in a fully automated manner, as in the case of BA determination, or in a semi-automated mode. Although BoneXpert, among other methods, was originally developed as AI-replace system, most radiologists are still looking at the hand radiographs to rule out abnormalities \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. The algorithm is based on the GP method for BA determination \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, which has been re-evaluated in different populations of the 21st century, indicating no advanced skeletal maturity due to secular change especially in Caucasian/European children \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. However, heterogeneities between and within different ethnic groups have been reported \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, and data show that the GP method should be used with caution in Asian and African populations \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. To address this limitation, the used software has been and is being validated across Asian, Hispanic, African and Caucasian populations. As there is a complex interrelationship between effects of socioeconomic status and ethnicity on BA estimation, there still remain uncertainties in increasingly heterogeneous populations as the modern central European one.\u003c/p\u003e \u003cp\u003eThe determination of bone age is important to properly assess and guide the evaluation and, if needed, therapy, of short or tall stature, impaired or accelerated growth, and delayed or early puberty. To the best of our knowledge, this study represents the first systematic evaluation of automated BA and BHI assessment in pLT patients. A significant positive correlation was observed between BA SDS and height percentiles. The tendency to have advanced BA in comparison to chronological age has been reported before in healthy tall children \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Patient age at LT was significantly negatively associated with the reduction in BA SDS post-LT, independent of whether the underlying condition was biliary atresia (in constituting a younger age group) or not. The basis for this observation remains to be clarified. Biliary atresia was the single most common indication for pLT in this study, consistent with data from larger study populations \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePediatric osteoporosis is defined as a clinically significant, low-energy fracture history (clinically significant fracture history is either of the following: \u0026ge;2 long-bone fractures by the age of 10 years or \u0026ge;\u0026thinsp;3 long-bone fractures at any age up to age 19 years) and a bone mineral density (BMD) Z-score \u0026le; \u0026minus;\u0026thinsp;2.0, or vertebral compression fractures independent of BMD \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. The diagnosis of pediatric osteoporosis should not be made based on densitometric criteria (i.e., DXA) alone. Besides ethnicity and genetics, there are complex and multifactorial influences on bone health in children and adolescents, i.e. nutrition and Vitamin D intake, exercise and lifestyle, body weight and composition aswell as hormonal status. In pLT patients, degree of cholestasis and immunosuppressive agents, especially glucocorticoids, can adversely affect bone metabolism \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. DXA can be useful to detect abnormal BMD and fracture risk in this population but comes with certain limitations in the pediatric population \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Furthermore, pediatric-experienced operators and child-specific interpretation are not universally available, and many centers lack pediatric-validated software and reference databases. The assessment of pediatric BMD remains an evolving field, as newer methods are being evaluated in children at risk for poor bone outcomes; among these, quantitative CT (QCT) is not routinely applied in children, but primarily utilized within the context of research studies \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. At present, no single imaging modality appears capable of providing a comprehensive assessment of bone health in children and adolescents, with each approach exhibiting its own limitations. According to Shalof et al., the digital X-ray radiogrammetry (DXR) method used for BHI assessment achieved the strongest positive relationship with DXA, compared to quantitative ultrasound and peripheral QCT, and should be further evaluated as a predictor of fractures in children and young adults \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. A significant positive correlation between BHI and BHI-SDS with the DXA and peripheral QCT readings was reported \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. DXR is observer-independent, inexpensive, widely available and offers the advantage of a very low radiation dose of \u0026lt;\u0026thinsp;0.00012 mSv per hand radiograph \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Automated adult height prediction based on BA assessment can be combined, contributing to its effectiveness in clinical practice.\u003c/p\u003e \u003cp\u003eReduced bone density is a common complication of chronic liver disease in both adults and children and relates to the presence of cholestasis \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. In this study, mean BHI SDS was reduced at all timepoints and notably, declined progressively until year 5 post-LT, independent of the underlying disease. This could indicate a reduced and further decreasing BMD in this cohort of pLT recipients. The quite proactive treatment approach of biliary complications including cholestatic disease post-LT could provide an explanation for why no difference exists between the groups with and without biliary complications. It should be noted that cholestasis cannot be assessed solely by radiological means and does not consistently manifest as visible dilatation of the bile ducts.\u003c/p\u003e \u003cp\u003eIn pathologic states like chronic kidney disease, higher PTH tends to correlate with poorer bone mineralisation \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. A previous study involving children with chronic kidney disease showed reduced BHI and a weakly significant negative correlation between BHI SDS and plasma PTH \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Within the present study, the correlation between BHI SDS and plasma PTH was found to be negative at year 1 (not significant) and 3 (weakly significant) post-LT; also, there was a negative correlation between BA SDS and plasma PTH post-LT. Notably, the relationship between PTH and bone density in pediatric chronic liver disease is not as straightforward as seen in adult hyperparathyroidism. Several pediatric studies do not find significantly elevated PTH in liver disease; deficits in BMD in these children seem to be more complex and multifactorial, and PTH cannot be considered as main correlate of low BMD in pediatric liver disease or pLT recipients.\u003c/p\u003e \u003cp\u003eAccording to previous data, the overall risk of sustaining a fracture during childhood is 10\u0026ndash;25% and the lifetime risk of sustaining a fracture 27% for girls and 42% for boys \u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. The relationship between fracture risk and BHI has been described in specific disease groups (like Duchenne muscle dystrophy). The small number of fractures in our cohort is likely underestimated within the framework of the retrospective data analysis. In this center, no screening for fractures had been carried out, as described elsewhere \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. In a recent study, an epidemiological association of BHI with both BMD and fracture risk was described, and the authors suggested that automated BHI assessment using hand radiographs might constitute a cost-effective measurement to identify children at risk of low BMD and fractures in pediatric clinics \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThis study has some limitations due to its retrospective and monocentric design. The study population is partially heterogeneous, including some rare underlying diseases. In addition, the small number of preoperative radiographs compared to postoperative ones considerably limited the comparison between these two time points. Although bone age and final height assessments are routinely performed in pLT recipients, a potential bias might exist due to the selection of patients who underwent radiographic examination. As DXA is not routinely performed in these patients, a comparison between the methods is not possible.\u003c/p\u003e \u003cp\u003eFurther studies with a longer follow-up period will be required to characterize bone health and assess fracture risk in pLT recipients.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eWhereas mean BA SDS remained approximately normotypic following pLT, mean BHI SDS was persistently reduced and exhibited a continuous decline through year 5. This may suggest a progressive decline in BMD post-LT. Further prospective studies with larger cohorts are needed to determine the utility of fully automated BA and BHI assessment in pLT recipients.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAI: artificial intelligence\u003c/p\u003e\n\u003cp\u003eAP: anterio-posterior\u003c/p\u003e\n\u003cp\u003eBA SDS: bone age standard deviation score\u003c/p\u003e\n\u003cp\u003eBHI SDS: bone health index standard deviation score\u003c/p\u003e\n\u003cp\u003eBMD: bone mineral density\u003c/p\u003e\n\u003cp\u003eCI: confidence interval\u003c/p\u003e\n\u003cp\u003eDXA: dual energy X-ray absorptiometry\u003c/p\u003e\n\u003cp\u003eDV: dorso-volar\u003c/p\u003e\n\u003cp\u003eGP: Greulich and Pyle\u0026nbsp;\u003c/p\u003e\n\u003cp\u003epLT: pediatric liver transplantation\u003c/p\u003e\n\u003cp\u003ePTCD: percutaneous transhepatic cholangiographic drainage\u003c/p\u003e\n\u003cp\u003eERCP: endoscopic retrograde cholangiopancreatography\u003c/p\u003e\n\u003cp\u003ePTH: parathyroid hormone\u003c/p\u003e\n\u003cp\u003eQCT: quantitative CT\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eStatement of contribution\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBG contributed to the research design, writing the paper, performance of the research and data analysis. All authors had full access to the data. GN contributed to data acquisition and writing the paper. FK contributed to data analysis. SH, QS, KB, CS contributed to revising the draft. GN contributed to the conception and design of the work and data analysis. SH contributed to writing the paper and final approval of the paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHuman Ethics and Consent to Participate declarations\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was performed in accordance with the ethical principles of the Declaration of Helsinki and its amendments and with the approval of the local IRB („Ethics Comittee, University Regensburg“; vote number: 24-4012-104).\u0026nbsp;Written informed consent was waived by the Institutional Review Board owing to the retrospective nature of the evaluation of radiographs acquired in routine clinical practice.\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\u003eDeclaration of Competing Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors declare no conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding source\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors received no financial support for the research, authorship and publication of this article. The scientific department of Bonexpert provided the software for retrospective analysis free of charge.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData sharing statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available on request from the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eVimalesvaran S, Verma A, Dhawan A. Pediatric Liver Transplantation: Selection Criteria and Post-transplant Medical Management. Indian J Pediatr. 2024;91(4):383\u0026ndash;90. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s12098-023-04963-5\u003c/span\u003e\u003cspan address=\"10.1007/s12098-023-04963-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaumann U, Karam V, Adam R, et al. Prognosis of Children Undergoing Liver Transplantation: A 30-Year European Study. Pediatrics. 2022;150(4). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1542/peds.2022-057424\u003c/span\u003e\u003cspan address=\"10.1542/peds.2022-057424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVimalesvaran S, Souza LN, Deheragoda M, et al. Outcomes of adults who received liver transplant as young children. EClinicalMedicine. 2021;38:100987. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.eclinm.2021.100987\u003c/span\u003e\u003cspan address=\"10.1016/j.eclinm.2021.100987\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKusumi K, Shaikhkhalil A, Patel HP, Mahan JD. Promoting bone health in children and adolescents following solid organ transplantation. Pediatr Transpl. 2021;25(1):e13940. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/petr.13940\u003c/span\u003e\u003cspan address=\"10.1111/petr.13940\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHelenius I, Remes V, Salminen S, et al. Incidence and predictors of fractures in children after solid organ transplantation: a 5-year prospective, population-based study. J Bone Min Res. 2006;21(3):380\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1359/JBMR.051107\u003c/span\u003e\u003cspan address=\"10.1359/JBMR.051107\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnastasilakis AD, Tsourdi E, Makras P, et al. Bone disease following solid organ transplantation: A narrative review and recommendations for management from The European Calcified Tissue Society. Bone. 2019;127:401\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bone.2019.07.006\u003c/span\u003e\u003cspan address=\"10.1016/j.bone.2019.07.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim WH, Ng CH, Ow ZGW, et al. A systematic review and meta-analysis on the incidence of osteoporosis and fractures after liver transplant. Transpl Int. 2021;34(6):1032\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/tri.13863\u003c/span\u003e\u003cspan address=\"10.1111/tri.13863\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCavallo F, Mohn A, Chiarelli F, Giannini C. Evaluation of Bone Age in Children: A Mini-Review. Front Pediatr. 2021;9:580314. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fped.2021.580314\u003c/span\u003e\u003cspan address=\"10.3389/fped.2021.580314\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin DD, Wit JM, Hochberg Z, et al. The use of bone age in clinical practice - part 1. Horm Res Paediatr. 2011;76(1):1\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1159/000329372\u003c/span\u003e\u003cspan address=\"10.1159/000329372\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin DD, Meister K, Schweizer R, Ranke MB, Thodberg HH, Binder G. Validation of automatic bone age rating in children with precocious and early puberty.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOffiah AC. Current and emerging artificial intelligence applications for pediatric musculoskeletal radiology. Pediatr Radiol. 2022;52(11):2149\u0026ndash;58. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00247-021-05130-8\u003c/span\u003e\u003cspan address=\"10.1007/s00247-021-05130-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin DD, Calder AD, Ranke MB, Binder G, Thodberg HH. Accuracy and self-validation of automated bone age determination. Sci Rep. 2022;12(1):6388. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-022-10292-y\u003c/span\u003e\u003cspan address=\"10.1038/s41598-022-10292-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin DD, Heckmann C, Jenni OG, Ranke MB, Binder G, Thodberg HH. Metacarpal thickness, width, length and medullary diameter in children\u0026mdash;reference curves from the First Z\u0026uuml;rich Longitudinal Study. Osteoporos Int. 2011;22(5):1525\u0026ndash;36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00198-010-1389-9\u003c/span\u003e\u003cspan address=\"10.1007/s00198-010-1389-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonnard GD. Cortical thickness and diaphysial diameter of the metacarpal bones from the age of three months to eleven years. Helv Paediatr Acta. 1968;23(5):445\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreen MA, Tsai A, Stamm A, Kleinman PK. Bone age assessment practices in infants and older children among Society for Pediatric Radiology members. Pediatr Radiol. 2016;46(9):1269\u0026ndash;74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00247-016-3618-7\u003c/span\u003e\u003cspan address=\"10.1007/s00247-016-3618-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ehttps://bonexpert.com. Accessed 13 January2026.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThodberg HH, Kreiborg S, Juul A, Pedersen KD. The BoneXpert method for automated determination of skeletal maturity. IEEE Trans Med Imaging. 2009;28(1):52\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TMI.2008.926067\u003c/span\u003e\u003cspan address=\"10.1109/TMI.2008.926067\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThodberg HH, van Rijn RR, Jenni OG, Martin DD. Automated determination of bone age from hand X-rays at the end of puberty and its applicability for age estimation. Int J Legal Med. 2017;131(3):771\u0026ndash;80. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00414-016-1471-8\u003c/span\u003e\u003cspan address=\"10.1007/s00414-016-1471-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThodberg HH, van Rijn RR, Tanaka T, Martin DD, Kreiborg S. A paediatric bone index derived by automated radiogrammetry. Osteoporos Int. 2010;21(8):1391\u0026ndash;400. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00198-009-1085-9\u003c/span\u003e\u003cspan address=\"10.1007/s00198-009-1085-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeubert K, Brunner E. A studentized permutation test for the non-parametric Behrens\u0026ndash;Fisher problem. Comput Stat Data Anal. 2007;51(10):5192\u0026ndash;204. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.csda.2006.05.024\u003c/span\u003e\u003cspan address=\"10.1016/j.csda.2006.05.024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKonietschke F, Placzek M, Schaarschmidt F, Hothorn LA. nparcomp: An R Software Package for Nonparametric Multiple Comparisons and Simultaneous Confidence Intervals. J Stat Soft. 2015;64(9):1\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18637/jss.v064.i09\u003c/span\u003e\u003cspan address=\"10.18637/jss.v064.i09\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatil I. Visualizations with statistical details: The 'ggstatsplot' approach. JOSS. 2021;6(61):3167. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.21105/joss.03167\u003c/span\u003e\u003cspan address=\"10.21105/joss.03167\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026uuml;decke D. ggeffects: Tidy Data Frames of Marginal Effects from Regression Models. JOSS. 2018;3(26):772. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.21105/joss.00772\u003c/span\u003e\u003cspan address=\"10.21105/joss.00772\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThodberg HH, Thodberg B, Ahlkvist J, Offiah AC. Autonomous artificial intelligence in pediatric radiology: the use and perception of BoneXpert for bone age assessment. Pediatr Radiol. 2022;52(7):1338\u0026ndash;46. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00247-022-05295-w\u003c/span\u003e\u003cspan address=\"10.1007/s00247-022-05295-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGREULICH WW, PYLE SI. RADIOGRAPHIC ATLAS OF SKELETAL DEVELOPMENT OF, THE HAND AND WRIST. Am J Med Sci. 1959;238(3):393. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/00000441-195909000-00030\u003c/span\u003e\u003cspan address=\"10.1097/00000441-195909000-00030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlshamrani K, Offiah AC. Applicability of two commonly used bone age assessment methods to twenty-first century UK children. Eur Radiol. 2020;30(1):504\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-019-06300-x\u003c/span\u003e\u003cspan address=\"10.1007/s00330-019-06300-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDahlberg PS, Mosd\u0026oslash;l A, Ding Y, et al. A systematic review of the agreement between chronological age and skeletal age based on the Greulich and Pyle atlas. Eur Radiol. 2019;29(6):2936\u0026ndash;48. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-018-5718-2\u003c/span\u003e\u003cspan address=\"10.1007/s00330-018-5718-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMansourvar M, Ismail MA, Raj RG, et al. The applicability of Greulich and Pyle atlas to assess skeletal age for four ethnic groups. J Forensic Leg Med. 2014;22:26\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jflm.2013.11.011\u003c/span\u003e\u003cspan address=\"10.1016/j.jflm.2013.11.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlshamrani K, Messina F, Offiah AC. Is the Greulich and Pyle atlas applicable to all ethnicities? A systematic review and meta-analysis. Eur Radiol. 2019;29(6):2910\u0026ndash;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-018-5792-5\u003c/span\u003e\u003cspan address=\"10.1007/s00330-018-5792-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJunge N, Karam V, Hartog H, et al. Update on pediatric liver transplantation in Europe 2022: An ELITA-ESPGHAN report. J Pediatr Gastroenterol Nutr. 2025;81(1):82\u0026ndash;90. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jpn3.70065\u003c/span\u003e\u003cspan address=\"10.1002/jpn3.70065\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeber DR, Boyce A, Gordon C, et al. The Utility of DXA Assessment at the Forearm, Proximal Femur, and Lateral Distal Femur, and Vertebral Fracture Assessment in the Pediatric Population: 2019 ISCD Official Position. J Clin Densitom. 2019;22(4):567\u0026ndash;89. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jocd.2019.07.002\u003c/span\u003e\u003cspan address=\"10.1016/j.jocd.2019.07.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJang MJ, Shin C, Kim S, et al. Factors affecting bone mineral density in children and adolescents with secondary osteoporosis. Ann Pediatr Endocrinol Metab. 2023;28(1):34\u0026ndash;41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.6065/apem.2244026.013\u003c/span\u003e\u003cspan address=\"10.6065/apem.2244026.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhalatbari H, Binkovitz LA, Parisi MT. Dual-energy X-ray absorptiometry bone densitometry in pediatrics: a practical review and update. Pediatr Radiol. 2021;51(1):25\u0026ndash;39. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00247-020-04756-4\u003c/span\u003e\u003cspan address=\"10.1007/s00247-020-04756-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShalof H, Dimitri P, Shuweihdi F, Offiah AC. Which skeletal imaging modality is best for assessing bone health in children and young adults compared to DXA? A systematic review and meta-analysis. Bone. 2021;150:116013. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bone.2021.116013\u003c/span\u003e\u003cspan address=\"10.1016/j.bone.2021.116013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSch\u0026uuml;ndeln MM, Marschke L, Bauer JJ, et al. A Piece of the Puzzle: The Bone Health Index of the BoneXpert Software Reflects Cortical Bone Mineral Density in Pediatric and Adolescent Patients. PLoS ONE. 2016;11(3):e0151936. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0151936\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0151936\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoomes KM, Spino C, Goodrich NP, et al. Bone Density in Children With Chronic Liver Disease Correlates With Growth and Cholestasis. Hepatology. 2019;69(1):245\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hep.30196\u003c/span\u003e\u003cspan address=\"10.1002/hep.30196\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLalayiannis AD, Crabtree NJ, Fewtrell M, et al. Assessing bone mineralisation in children with chronic kidney disease: what clinical and research tools are available? Pediatr Nephrol. 2020;35(6):937\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00467-019-04271-1\u003c/span\u003e\u003cspan address=\"10.1007/s00467-019-04271-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eN\u0026uuml;sken E, Imschinetzki D, N\u0026uuml;sken K-D, et al. Automated Greulich-Pyle bone age determination in children with chronic kidney disease. Pediatr Nephrol. 2015;30(7):1173\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00467-015-3042-5\u003c/span\u003e\u003cspan address=\"10.1007/s00467-015-3042-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarsen AV, Mundbjerg E, Lauritsen JM, Faergemann C. Development of the annual incidence rate of fracture in children 1980\u0026ndash;2018: a population-based study of 32,375 fractures. Acta Orthop. 2020;91(5):593\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/17453674.2020.1772555\u003c/span\u003e\u003cspan address=\"10.1080/17453674.2020.1772555\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrijatelj V, Grgic O, Uitterlinden AG, Wolvius EB, Rivadeneira F, Medina-Gomez C. Bone health index in the assessment of bone health: The Generation R Study. Bone. 2024;182:117070. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bone.2024.117070\u003c/span\u003e\u003cspan address=\"10.1016/j.bone.2024.117070\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmim","sideBox":"Learn more about [BMC Medical Imaging](http://bmcmedimaging.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bmim/default.aspx","title":"BMC Medical Imaging","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Child/adolescent, liver transplantation, bone age measurement, radiography, hand, bone density","lastPublishedDoi":"10.21203/rs.3.rs-8694835/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8694835/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e \u003cp\u003eBone metabolism in children who have undergone pediatric liver transplantation (pLT) can be negatively affected, particularly in the presence of cholestasis. Pediatric bone status is a critical determinant of bone health through adulthood. The aim of this study was to evaluate bone age (BA), a marker of skeletal maturity, and bone health index (BHI), a surrogate marker of bone density, in pLT recipients.\u003c/p\u003e\u003ch2\u003eMethods:\u003c/h2\u003e \u003cp\u003e A total of 449 left hand radiographs of 207 patients [median age at LT: 1.1 years (IQR: 0.43; 6.8 years); female\u0026thinsp;=\u0026thinsp;50.2%], all entered in the local pLT register, were evaluated in this retrospective, IRB-approved single-center exploratory study. Fully automated determination of BA-standard deviation score (BA SDS) and BHI SDS was performed using a commercial, CE-marked AI tool and relationships with age, sex, underlying disease (biliary atresia vs. non-biliary atresia), height percentile and plasma parathyroid hormone (PTH) were assessed using t-tests, permutation tests, and Pearson correlation in both the whole cohort and the subgroup with biliary complications.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e \u003cp\u003eMean BA SDS pre-LT was 0.03\u0026thinsp;\u0026plusmn;\u0026thinsp;1.50, 0.03\u0026thinsp;\u0026plusmn;\u0026thinsp;1.70 at year 1, -0.07\u0026thinsp;\u0026plusmn;\u0026thinsp;1.62 at year 3 and \u0026minus;\u0026thinsp;0.21\u0026thinsp;\u0026plusmn;\u0026thinsp;1.63 at year 5. Mean BHI SDS pre-LT was \u0026minus;\u0026thinsp;1.8\u0026thinsp;\u0026plusmn;\u0026thinsp;1.20, -1.6\u0026thinsp;\u0026plusmn;\u0026thinsp;1.20 at year 1, -2.0\u0026thinsp;\u0026plusmn;\u0026thinsp;1.10 year 3 and \u0026minus;\u0026thinsp;2.2\u0026thinsp;\u0026plusmn;\u0026thinsp;1.3 at year 5. A significant positive correlation was observed between BA SDS and height percentiles at all timepoints. PTH showed a significant inverse correlation with BA SDS at years 3 and 5 and with BHI SDS at year 5. No significant differences in mean BA SDS or BHI SDS were observed between groups with or without biliary complications or PTCD/ERCP at any timepoint.\u003c/p\u003e\u003ch2\u003eConclusions:\u003c/h2\u003e \u003cp\u003eWhile mean BA SDS remained stable around zero post-LT, mean BHI SDS was consistently reduced and declined progressively until year 5, indicating decreased bone mineral density. Further prospective studies with larger cohorts are needed to determine the utility of automated BA and BHI assessment following pLT.\u003c/p\u003e","manuscriptTitle":"Automated determination of bone age and bone health index in pediatric liver transplant recipients","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-11 12:03:00","doi":"10.21203/rs.3.rs-8694835/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-03-06T05:34:01+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-04T06:30:37+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-02-05T04:36:44+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-04T20:41:12+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Imaging","date":"2026-02-04T20:34:58+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmim","sideBox":"Learn more about [BMC Medical Imaging](http://bmcmedimaging.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bmim/default.aspx","title":"BMC Medical Imaging","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2a719b41-96ed-4d81-a2e8-b0a716ebab3f","owner":[],"postedDate":"March 11th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-11T12:03:01+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-11 12:03:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8694835","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8694835","identity":"rs-8694835","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}