Automatic bone age assessment for adult height prediction in children with congenital adrenal hyperplasia.

Predicted adult height (PAH) can be calculated using methods such as Bayley-Pinneau (BP), Roche-Wainer-Thissen (RWT), and BoneXpert based on bone age (BA) assessment. Since these methods were developed for healthy children, varying results have been reported regarding their efficacy across different patient groups. This study aimed to determine the most effective method for PAH by comparing the BP, RWT, and BoneXpert methods in boys with constitutional delay of growth and puberty (CDGP). Sixty-two male patients with CDGP who had reached their final height (FH) were included in the study. Two experienced clinicians reassessed left-hand and wrist radiographs taken at the time of diagnosis using the Greulich-Pyle (GP) atlas to manually determine BA. Among the methods used for PAH, the GP atlas was used for BP and RWT, while we used its GP-based electronic software for the BoneXpert method. The mean age at diagnosis of the patients was 14.2 ± 0.8 years, with 58.1% (n=36) having a similar family history. The mean height standard deviation (SD) score was -2.1 ± 0.9, and 24.2% (n=15) of patients received low-dose testosterone induction therapy. The median BAs of the patients were 12.5 (11.5-13.0) years using the GP atlas and 12.6 (11.8-13.4) years with BoneXpert (p<0.001). Boys who were treated with or not treated with testosterone therapy had similar mean height SD scores, median testicular volumes, and median BAs assessed by both methods. The mean target height and FH SD scores were -0.6 ± 0.6 and -0.6 ± 0.9, respectively (p=0.8). Almost all patients (n=60, 97%) achieved adult height within the target range, with no significant difference in the FH SD score between boys who received testosterone and those who did not (p=0.1). There was no significant difference between the FH and PAH when estimated by the BP and RWT methods (p=0.2 and p=0.6, respectively), while the BoneXpert method underestimated the FH (p<0.001). The BP and RWT methods provided better predictions in patients with BA ≤ 2 years compared to BoneXpert (p=0.3 and p=0.4 vs. p<0.001, respectively). Conversely, RWT and BoneXpert methods were more accurate in PAH in boys with delayed BA >2 years (p=0.1 and p=0.1, respectively), while the BP method resulted in overestimation (p=0.003). The RWT method was found to be a better predictor of FH compared to the BP or BoneXpert methods in boys with delayed BA ≤ 2 years and >2 years.

Background: Bone age assessments (BAAs) is an important clinical modality to investigate endocrine, genetic and growth disorders in children. It is generally performed by radiological examination of the left hand by using either the Greulich-Pyle (GP) or the Tanner-Whitehouse (TW) method. However, both clinical procedures show several limitations, from significant intra- and inter-operator variability to examination effort of clinicians. To address these problems, several automated approaches have been proposed; nevertheless, some disparity still exists between automated BAAs and manual BAAs to be employed in clinical practice. To overcome this disparity, deep learning-based bone age assess software using GP and TW3 hybrid method has been developed. In this study, we evaluate the accuracy and efficiency of the new automated hybrid software system for bone age assessment and validate its feasibility in clinical practice. Materials and Methods: Greulich-Pyle (GP) and Tanner-Whitehouse (TW3) hybrid method-based deep-learning technique was used to develop the automated software system for bone age assessment. Total 102 radiographs from children with the chronological age of 4.9-17.0 years (mean age 10.9±2.3, 51 cases for females and 51 cases for males) were selected and bone age was estimated with this software. For validation of the automated software system, three human experts have manually performed BAAs at expert’s discretion based on GP method for accuracy estimation and one naïve radiologist performed BAAs with automated software system assist and BAAs reading time was recorded in each session for efficiency evaluation. The performance of automated software system was assessed by comparing mean absolute difference (MAD) between the system estimates and the experts manual BAAs.Results: The results of bone age assessment by human experts and automated software system showed no significant difference between the two groups. Each assessed average of bone age were 11.39 ± 2.74 and 11.35 ± 2.76, respectively. MAD was 0.39 years between automated software system BAAs and experts manual BAAs. The 95% confidence interval of the MAD was 0.33 years and 0.45 years. BAAs reading time was reduced from 56.81 sec (95% confidence interval 52.81 - 60.81 sec) in naïve manual BAAs to 31.72 sec (95% confidence interval 29.74 - 33.69 sec) in automated software system assisted BAAs and statistically significant (p < 0.001). MAD showed 0.42 years between naïve manual BAAs and the software-assisted BAAs (95% confidence interval 0.31-0.47 years).Conclusion: The newly developed GP and TW3 hybrid automated software system were reliable for bone age assessments with equivalent accuracy to human experts. Also, the automated system appeared to enhance efficiency by reducing reading times without compromising diagnostic accuracy.

Point-of-Care Bone Age Evaluation: The Increasing Role of US in Resource-limited Populations

Back to basics: Early diagnosis and compliance improve final height outcome in congenital adrenal hyperplasia

IntroductionAutomated bone age assessment has recently become increasingly popular. The aim of this study was to assess the agreement between automated and manual evaluation of bone age using the method according to Tanner-Whitehouse (TW3) and Greulich-Pyle (GP).MethodsWe evaluated 1285 bone age scans from 1202 children (657 scans from 612 boys) by using both manual and automated (TW3 as well as GP) bone age assessment. BoneXpert software versions 2.4.5.1. (BX2) and 3.2.1. (BX3) (Visiana, Holte, Denmark) were compared with manual evaluation using root mean squared error (RMSE) analysis.ResultsRMSE for BX2 was 0.57 and 0.55 years in boys and 0.72 and 0.59 years in girls, respectively for TW3 and GP. For BX3, RMSE was 0.51 and 0.68 years in boys and 0.49 and 0.52 years in girls, respectively for TW3 and GP. Sex- and age-specific analysis for BX2 identified the largest differences between manual and automated TW3 evaluation in girls between 6-7, 12-13, 13-14 and 14-15 years, with RMSE 0.88, 0.81, 0.92 and 0.84 years, respectively. The BX3 version showed better agreement with manual TW3 evaluation (RMSE 0.64, 0.45, 0.46 and 0.57).ConclusionThe latest version of the BoneXpert software provides improved and clinically sufficient agreement with manual bone age evaluation in children of both sexes compared to the previous version and may be used for routine bone age evaluation in non-selected cases in pediatric endocrinology care.

ObjectiveTo evaluate the accuracy and clinical efficacy of a hybrid Greulich-Pyle (GP) and modified Tanner-Whitehouse (TW) artificial intelligence (AI) model for bone age assessment.Materials and MethodsA deep learning-based model was trained on an open dataset of multiple ethnicities. A total of 102 hand radiographs (51 male and 51 female; mean age ± standard deviation = 10.95 ± 2.37 years) from a single institution were selected for external validation. Three human experts performed bone age assessments based on the GP atlas to develop a reference standard. Two study radiologists performed bone age assessments with and without AI model assistance in two separate sessions, for which the reading time was recorded. The performance of the AI software was assessed by comparing the mean absolute difference between the AI-calculated bone age and the reference standard. The reading time was compared between reading with and without AI using a paired t test. Furthermore, the reliability between the two study radiologists' bone age assessments was assessed using intraclass correlation coefficients (ICCs), and the results were compared between reading with and without AI.ResultsThe bone ages assessed by the experts and the AI model were not significantly different (11.39 ± 2.74 years and 11.35 ± 2.76 years, respectively, p = 0.31). The mean absolute difference was 0.39 years (95% confidence interval, 0.33–0.45 years) between the automated AI assessment and the reference standard. The mean reading time of the two study radiologists was reduced from 54.29 to 35.37 seconds with AI model assistance (p < 0.001). The ICC of the two study radiologists slightly increased with AI model assistance (from 0.945 to 0.990).ConclusionThe proposed AI model was accurate for assessing bone age. Furthermore, this model appeared to enhance the clinical efficacy by reducing the reading time and improving the inter-observer reliability.

The purpose of this investigation was to develop a quantitative and reproductible method for estimating skeletal maturity based on measurements of the height of the bony greater trochanter (GT) using timing to 90% of final height as a gold standard. Bony GT height was measured using serial anteroposterior pelvic radiographs in 76 healthy pediatric patients obtained from the Bolton-Brush (BB) Study with corresponding Greulich-Pyle (GP) bone ages. Chronologic age at 90% of final height was calculated. GT height was then measured in 300 contemporary patients aged 4 to 18 years, evenly divided based on sex and race. Bony GT height was compared between BB and contemporary patients, while linear mixed-effects models were used to examine for potential predictors of years to 90% final height using patient sex, GP bone age and bony GT height measurements. Bony GT height was measured in 303 radiographs from the BB Collection (n=37 males; n=39 females) with corresponding GP bone ages, chronological ages, and heights to represent skeletal maturity. Mean age at 90% final height was 13.3±0.6 years for males and 11.4±0.8 years for females. When controlling for patient age and sex, multiple regression analysis revealed that contemporary patients possessed significantly greater bony GT height (mean difference: 1.15 mm; P=0.001) when compared with BB patients. Multivariate analysis showed that combining bony GT height, GP bone age, and sex significantly predicted years to 90% total growth (P<0.001) and explained ∼85% (95% confidence interval for R2: 82%-87%) of the total variance in years using 90% of final height, with sex, GP bone age, and GT height all significant contributors. Including bony GT height provides more accurate prediction of 90% final height when combined with GP bone age and sex. GT height offers an efficient and accurate parameter that may be utilized in pediatric orthopedic conditions requiring a quantitative estimate of bone age in children with prior pelvis or hip imaging. Level II-diagnostic study.

Bone age assessment (BAA) using radiological x-rays of the left-hand wrist is important in pediatric endocrinology to correctly assess growth and pubertal maturation. To identify the age, the most commonly used technique is bone age assessment. The bone age is the most commonly used criteria in age and growth disorder. There are two main methods for bone age estimation such as Tanner & Whitehouse (TW) method and Greulich & Pyle (GP) method. Tanner & Whitehouse (TW) method is a score assigning method while Greulich & Pyle (GP) is an atlas matching method. The objective of this research is to improve the accuracy of the automated bone age estimation. To improve the robustness and accuracy of this system, image preprocessing techniques are also necessary. Bone age assessment can affect the accuracy in segmentation because of poor contrast, noise, and various hand positions. Preprocessing steps of this system include background removal, radiological marker removal, image enhancement, and hand rotation methods. In this paper, we use the Contrast Limited Adaptive Histogram Equalization (CLAHE) method to enhance the contrast of the hand bone radiograph. This method enhances the hand radiograph in which the background and body have more brightness. The advantage of CLAHE is to prevent the over-amplification of noise, it is able to increase contrast. In this paper, we tested contrast enhancement techniques such as Contrast Limited Adaptive Histogram Equalization (CLAHE), Histogram Equalization (HE) and Power Law Transform (PLT) techniques.

Final height (FH) in congenital adrenal hyperplasia (CAH) patients is frequently under -1.5 SDS in relation to the normal population, even in those with good compliance, chronic androgen exposure or supraphysiological glucocorticoid doses could corroborate to this finding. Recombinant human growth hormone (GH) has been used as adjuvant therapy to improve FH in many disorders, but there are scarce data in CAH patients.Objectiveto analyze the effect of GH therapy on final height of CAH patients with compromised height prediction.Patients and Methodsdata of 26 patients (7F/19M) were retrospectively evaluated: 10 salt wasting-SW, 11 simple virilizing-SV and 5 with nonclassical form-NC. The SW patients were diagnosed at the neonatal period, SV and NC patients at mean chronological age of 3.6±2 and 7.2±3.8 yrs, respectively. All patients received cortisone acetate, hydrocortisone equivalent dose of 12.6±6.6 mg/m2/day in classical and 8.24±4.7 mg/m2/day in NC form; SW patients also received fludrocortisone 36±9.6mcg/m2/day. Near final height (NFH) was considered as BA of 13yrs for girls and 15yrs for boys; height and target height (TH) were described as SDS. All patients presented hormonal/molecular CAH diagnosis. The t test was used in the statistical analysis.ResultsMean chronological and bone age (BA) at the beginning of GH therapy was 9.0±2.4 and 12.3±2yrs, respectively, mean duration of GH therapy was 4.1±1.9yrs and daily dose was 0.15 UI/kg/day. All patients were submitted to pubertal blockage with leuprorelin acetate 11.25 mg each 3 months during 3.9±1.1yrs or cyproterone acetate, mean dose 70±19 mg/m2/d during 4.7±2.5yrs. At the last evaluation, 19 patients (14M/5F) achieved FH: at the beginning of GH therapy mean H/BA in males was -2.1±0.38 and after -0.65±1 SDS (p<0.001). Mean FH in males was 165.7±5.9 cm and mean TH 170.7±8 cm (p>0.05); the TH-FH was -0.6±1.2 SDS. Among females, mean H/BA at the beginning of GH therapy was -2.2±0.7 and after -0.84±0.9 SDS (p=0.01). Mean FH in females was 156±8 cm, mean TH was 160±5.8cm (p>0.05); the TH-FH was -0.6±0.7 SDS. Considering patients with FH and NFH, the FH-TH in males was 0.4±0.2 SDS and -0.15±0.9 SDS in females. An improvement of approximately 12.7cm in males and 12.4cm in females was observed. No patient presented hypertension and 7/26 patients presented insulin resistance, which normalized after GH withdrawal.Conclusionwe observed that GH therapy associated with puberty blockers allowed the achievement of final height in the expected range of target height, without significant side effects.Presentation: Saturday, June 11, 2022 1:00 p.m. - 3:00 p.m., Sunday, June 12, 2022 1:00 p.m. - 1:05 p.m.

In cases of precocious puberty, the determination of bone age (BA) is usually performed by clinicians using the Greulich-Pyle (GP) atlas, and there can be significant variation between assessors. The aim of this study was to compare predicted adult height (PAH) calculations based on BA read by the automated BA method “BoneXpert” (BX) with clinician-determined BA-based PAH calculations. Girls who presented with suspicion of precocious puberty and normal pubertal variants, such as premature thelarche and premature adrenarche, and whose BA determined by both BX and two different clinicians were followed up until reaching near-final height (NFH). Those whose breast development started before the age of 8 years were considered as precocious puberty. Four PAH calculations were performed with two different estimated height calculation methods, the Bayley-Pinneau (BP) and Roche-Wainer-Thissen based on two different BA predictions (Clinician-GP and BX-GP). PAH-standard deviation score (PAH-SDS) and NFH-SDS values of the patients were compared. The median chronological age of the 44 girls included at presentation was 9.3 years, while the median BA was 10.4 years and 10.6 years according to clinician-GP and BX-GP, respectively; mean height-SDS was 0.75 and target height-SDS was -0.28. When they reached NFH, the height-SDS was -0.02. Final analyzes were performed in 26 cases who did not have low birth weight and did not receive puberty-arresting treatment. Delta PAH-SDS-NFH-SDS (Δ-SDS) was compared according to the four different PAH calculations. The closest PAH-SDS value measurement to NFH-SDS was calculated by BP based on BA determined by the BX-GP method (-0.09). PAH calculations using the BP method based on BX-derived GP readings most accurately predict NFH in girls with precocious puberty, and normal pubertal variants.

The aim of this study was to identify and compare bone age assessments of Turkish children in regions of Central Anatolia and Eastern Anatolia using the Greulich-Pyle (GP) method. Evaluation was made of 849 (375 boys, 514 girls) adolescents aged between 9 and 17 years, who lived in two geographically different regions of Turkey. The selection criteria included normal growth and development, a state of good physical and mental health with no previous history of chronic or acute illnesses, no past trauma or injury to the hand-wrist region, no congenital or acquired malformations of the hand-wrist area, no hormonal disorders, and good quality hand-wrist radiographs. Bone age (BA) was evaluated using the GP method from definitive radiographs of the left hand-wrist. The total mean differences between BA and chronological age (CA) for girls and boys were found to be 1.19 ± 1.2 (p < 0.05) and 0.90 ± 1.2 (p < 0.05) years in the Eastern Anatolia region and 0.28 ± 0.6 (p > 0.05) and –0.10 ± 0.3 (p > 0.05) years in the Central Anatolia region. There were significant differences between BA and CA in all age groups for girls and boys in Malatya (a city in the Eastern Anatolia region) and in 10 and 13 year olds for girls and 12, 13, 15, and 16 year olds for boys in Sivas (a city in the Central Anatolia region). Statistically significant differences were determined between the BA and CA of the subjects living in Malatya and Sivas (p < 0.05). The results of this study using the Greulich–Pyle atlas indicate that bone development is completed earlier in adolescents living in Malatya compared with those in Sivas.

Management and outcome of central precocious puberty

To develop and evaluate a simplified method for skeletal age assessment in adolescents as a rapid alternative to the Greulich-Pyle (GP) method. Retrospective methodological study comparing the simplified method with the GP atlas, with duplicate readings by experienced and inexperienced raters. One hundred and seventeen hand and wrist radiographs (67 boys, 50 girls) from adolescents referred for short stature but otherwise healthy. Radiographs with GP-assessed bone age ≥ 9 years in girls and ≥ 11 years in boys were included. Bone age was scored independently using GP and simplified methods in two rounds by two paediatric endocrinologists and, for the simplified method, two residents. Outcomes included measures of agreement between methods, the proportion of readings within ±1 year of GP, and inter- and intra-rater reliability. Among experienced raters, the simplified method showed minimal bias relative to GP, with about 91% of readings within ±1 year. Reliability coefficients for both methods exceeded 0.9. Repeatability of the simplified method was slightly lower than GP but remained within clinically acceptable limits and close to predefined acceptability thresholds for inexperienced raters. The simplified method offers a transparent, standardised, and reproducible approach to adolescent skeletal age assessment. Although not intended to replace the GP method, it is particularly suited to research involving large datasets or resource-limited environments and to population-level skeletal analyses. In busy clinics, it may also serve as a quick check of radiologists' GP-based interpretations, allowing review of reported bone ages without repeating a full atlas assessment.

Titration of growth hormone dose using insulin-like growth factor-1 measurements: Is it feasible in children?

Skeletal bone age assessment using X-ray images is a standard clinical procedure to detect any anomaly in bone growth among kids and babies. The assessed bone age indicates the actual level of growth, whereby a large discrepancy between the assessed and chronological age might point to a growth disorder. Hence, skeletal bone age assessment is used to screen the possibility of growth abnormalities, genetic problems, and endocrine disorders. Usually, the manual screening is assessed through X-ray images of the non-dominant hand using the Greulich–Pyle (GP) or Tanner–Whitehouse (TW) approach. The GP uses a standard hand atlas, which will be the reference point to predict the bone age of a patient, while the TW uses a scoring mechanism to assess the bone age using several regions of interest information. However, both approaches are heavily dependent on individual domain knowledge and expertise, which is prone to high bias in inter and intra-observer results. Hence, an automated bone age assessment system, which is referred to as Attention-Xception Network (AXNet) is proposed to automatically predict the bone age accurately. The proposed AXNet consists of two parts, which are image normalization and bone age regression modules. The image normalization module will transform each X-ray image into a standardized form so that the regressor network can be trained using better input images. This module will first extract the hand region from the background, which is then rotated to an upright position using the angle calculated from the four key-points of interest. Then, the masked and rotated hand image will be aligned such that it will be positioned in the middle of the image. Both of the masked and rotated images will be obtained through existing state-of-the-art deep learning methods. The last module will then predict the bone age through the Attention-Xception network that incorporates multiple layers of spatial-attention mechanism to emphasize the important features for more accurate bone age prediction. From the experimental results, the proposed AXNet achieves the lowest mean absolute error and mean squared error of 7.699 months and 108.869 months2, respectively. Therefore, the proposed AXNet has demonstrated its potential for practical clinical use with an error of less than one year to assist the experts or radiologists in evaluating the bone age objectively.