Charting the growth through intelligence: A SWOC analysis on AI-assisted radiologic bone age estimation.

Assessing Bone Age: A Paradigm for the Next Generation of Artificial Intelligence in Radiology.

This narrative review focuses on clinical applications of artificial intelligence (AI) in musculoskeletal imaging. A range of musculoskeletal disorders are discussed using a clinical-based approach, including trauma, bone age estimation, osteoarthritis, bone and soft-tissue tumors, and orthopedic implant-related pathology. Several AI algorithms have been applied to fracture detection and classification, which are potentially helpful tools for radiologists and clinicians. In bone age assessment, AI methods have been applied to assist radiologists by automatizing workflow, thus reducing workload and inter-observer variability. AI may potentially aid radiologists in identifying and grading abnormal findings of osteoarthritis as well as predicting the onset or progression of this disease. Either alone or combined with radiomics, AI algorithms may potentially improve diagnosis and outcome prediction of bone and soft-tissue tumors. Finally, information regarding appropriate positioning of orthopedic implants and related complications may be obtained using AI algorithms. In conclusion, rather than replacing radiologists, the use of AI should instead help them to optimize workflow, augment diagnostic performance, and keep up with ever-increasing workload.Relevance statement This narrative review provides an overview of AI applications in musculoskeletal imaging. As the number of AI technologies continues to increase, it will be crucial for radiologists to play a role in their selection and application as well as to fully understand their potential value in clinical practice.Key points• AI may potentially assist musculoskeletal radiologists in several interpretative tasks.• AI applications to trauma, age estimation, osteoarthritis, tumors, and orthopedic implants are discussed.• AI should help radiologists to optimize workflow and augment diagnostic performance.Graphical

Background and Objectives: Artificial intelligence (AI) has seen rapid integration into various areas of medicine, particularly with the advancement of machine learning (ML) and deep learning (DL) techniques. In pediatric orthopedics, the adoption of AI technologies is emerging but still not comprehensively reviewed. The purpose of this study is to review the latest evidence on the applications of artificial intelligence in the field of pediatric orthopedics. Materials and Methods: A literature search was conducted using PubMed and Web of Science databases to identify peer-reviewed studies published up to March 2024. Studies involving AI applications in pediatric orthopedic conditions—including spinal deformities, hip disorders, trauma, bone age assessment, and limb discrepancies—were selected. Eligible articles were screened and categorized based on application domains, AI models used, datasets, and reported outcomes. Results: AI has been successfully applied across several pediatric orthopedic subspecialties. In spinal deformities, models such as support vector machines and convolutional neural networks achieved over 90% accuracy in classification and curve prediction. For developmental dysplasia of the hip, deep learning algorithms demonstrated high diagnostic performance in radiographic interpretation. In trauma care, object detection models like YOLO and ResNet-based classifiers showed excellent sensitivity and specificity in pediatric fracture detection. Bone age estimation using DL models often matched or outperformed traditional methods. However, most studies lacked external validation, and many relied on small or single-institution datasets. Concerns were also raised about image quality, data heterogeneity, and clinical integration. Conclusions: AI holds significant potential to enhance diagnostic accuracy and decision making in pediatric orthopedics. Nevertheless, current research is limited by methodological inconsistencies and a lack of standardized validation protocols. Future efforts should focus on multicenter data collection, prospective validation, and interdisciplinary collaboration to ensure safe and effective clinical integration.

Bone age (BA) reflects skeletal maturity and is crucial in clinical and forensic contexts, particularly for growth assessment, adult height prediction, and managing conditions like short stature and precocious puberty, often using X-ray, MRI, CT, or ultrasound imaging. Traditional BA assessment methods, including the Greulich-Pyle and Tanner-Whitehouse techniques, compare morphological changes to reference atlases. Despite their effectiveness, factors like genetics and environment complicate evaluations, emphasizing the need for new methods that account for comprehensive variations in skeletal maturity. The limitations of classical BA assessment methods increase the demand for automated solutions. The first automated tool, HANDX, was introduced in 1989. Researchers now focus on developing reliable artificial intelligence (AI)-driven tools, utilizing machine learning and deep learning techniques to improve accuracy and efficiency in BA evaluations, addressing traditional methods' shortcomings. Recent reviews on BA assessment methods rarely compare AI-based approaches across imaging technologies. This article explores advancements in BA estimation, focusing on machine learning methods and their clinical implications while providing a historical context and highlighting each approach's benefits and limitations.

Bone age assessments were related to auxological variables in 407 Italian boys, between 7 and 12 years of age, in order to elucidate the factors that affect the rate of skeletal maturation and to examine the possibility of using measures of skeletal maturation of evaluate individual patients. Using the radius-ulna-short bones (RUS) method of assessment, bone age velocity was greater in the Italian boys than for the UK reference standards, although there was considerable interindividual dispersion around the mean. Bone age velocity and height velocity were poorly correlated, and there was little correlation between skeletal and pubertal maturation. There was a slight positive correlation between bone age velocity and height SDS and between bone age velocity and body mass index. Bone age estimations using RUS were greater than those obtained using the carpus. In conclusion, the marked interindividual deviation in measured bone ages makes it difficult to relate data on an individual basis to other measures of growth and maturation

Background Bone age assessment represents an important step in the management of children with Isolated Growth Hormone Deficiency (IGHD). This study examined the usefulness of Ultrasound (US) in the assessment of bone age in a sample of Iraqi children with IGHD as compared to radiography as a reference. Additionally, it verified if patient gender and growth hormone therapy have an impact on US accuracy. Methods An observational cross-sectional study recruited children with isolated growth hormone deficiency who were diagnosed and followed at the Alresafa Specialized Center for Endocrinology and Diabetes, Baghdad, Iraq, over 6 months. Children with IGHD from Iraqi nationality were recruited, while children from other nationalities or having multiple hormonal deficiencies, syndromic features, and parent refused participation were excluded. For each patient, a bone age assessment was conducted using two methods: US and TW2 hand-wrist radiographs at the same visit by the same radiologist. Results A total of 116 children were included. The chronological age of recruited children was 7 to 17 years, with a mean of 13.01 ± 2.78 years. There were 67 males (57.9%) with a male-to-female ratio of 1.37:1. The patient's gender did not affect the US accuracy; there was a non-significant difference in the bone age estimated by the US and conventional radiograph for both male and female patients, (p-value = 0.087, 0.308) respectively. Those who received growth hormone therapy and those who did not for both male and female patients (p-value = 0.071,0.243), respectively. There was a strong positive correlation between the means of bone age assessed by ultrasound (US) and conventional radiography for both males and females, with correlation coefficients of r = 0.788 and r = 0.703, respectively. Conclusion Ultrasound may serve as a valid replacement for radiography in the assessment of bone age in children with short stature caused by a growth hormone deficiency, irrespective of the gender and treatment received. Thus, it may overcome radiography drawbacks for children who need sequential bone age assessment.

OBJECTIVE: The skeletal maturity of any individual is known as bone age and it can be reliably estimated by a roentgenologic study of osseous development since the appearance and union of the centers of ossification occur in a fairly definite pattern and time sequence from birth to maturity. This is particularly helpful in the clinical workup of children with endocrinological disorders where skeletal and pubertal growth is affected. Therefore the bone age can be advanced, delayed or appropriate for chronological age in various endocrinological disorders. We present cases which affect skeletal maturation in a way that a simple investigation like the bone age estimation can lead to their diagnosis or can help therapeutic interventions. The present study aims to reiterate the importance of bone age in the diagnosis and management of endocrinological disorders. MATERIAL AND METHODS: X rays of both hands of cases and for bone age estimation we used the Greulich - Pyle atlas to accord a bone age to each bone of the hand and obtain an average reading. RESULT: The endocrinological disorders were classified as advanced, delayed and bone age appropriate for chronological age. Cases with advanced bone age were a child with hypothalamic hamartoma and a child with congenital adrenal hyperplasia. Those with delayed bone age were hypothyroidism and growth hormone deficiency. The case with bone age appropriate for chronological age was that of a Turner girl. In each of these cases the bone age helped in the diagnosis and also in the therapeutic decisions. CONCLUSIONS: Assessing bone age through radiographs of both hands graded according to Greulich Pyle method provided the endocrinologists an inexpensive tool to obtain an objective measure of the child's developmental status and also an effective tool for the diagnosis and differential diagnosis of various endocrinological disorders. It also guides the endocrinologist of the timing and effectiveness of the therapeutic measures employed.

Chapter 10 - Bone age assessment using metric learning on small dataset of hand radiographs

Background Bone age assessment is a critical tool in pediatric orthopedic surgery, guiding treatment decisions for growth-related disorders and surgical interventions. Traditional methods, such as the Greulich-Pyle and Tanner-Whitehouse techniques, rely on manual interpretation of hand and wrist radiographs, making them time-intensive and susceptible to inter-operator variability. Artificial intelligence (AI) has emerged as a promising tool to enhance accuracy, efficiency, and standardization in skeletal maturity assessment. Methods This study evaluates the application of AI in pediatric bone age prediction using the Radiological Society of North America (RSNA) 2017 Pediatric Bone Age Challenge dataset. A deep learning model based on the ResNet-50 architecture(Microsoft Research, Redmond, Washington, USA) was developed and trained on 12,611 hand and wrist radiographs, validated on 1,425 images, and tested on 200 images. Model performance was assessed using root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). Results The AI model achieved an RMSE of 11.07 months, an MAE of 8.54 months, and an R² of 0.929, indicating strong alignment with radiologist-determined bone ages. The Pearson correlation coefficient (0.963) and Spearman's rank correlation (0.955) confirmed the model's predictive robustness. Compared to traditional methods, which have reported variability with errors ranging from 6 to 18 months, the AI model demonstrated a reduction in inter-operator variability and improved reliability. Conclusion The implementation of AI in bone age assessment offers a more standardized, rapid, and precise alternative to conventional methods. By improving the accuracy and efficiency of skeletal maturity evaluations, AI has significant implications for pediatric orthopedic surgery, optimizing treatment timing and expanding access to high-quality bone age assessments. Further validation studies are needed to ensure clinical applicability across diverse patient populations.

Background:There are several methods of bone age (BA) assessment, which include Gruelich-Pyle (GP), Gilsanz-Ratib (GR), and Tanner Whitehouse-3 (TW-3) methods. Although GP atlas is the most widely used, there are concerns about its accuracy in children of different ethnicities, making the use of the TW-3 method an attractive option in Indian children.Objectives:1) To assess the relationship of BA with chronological age (CA) as assessed by different methods (GP, GR, and TW-3) in healthy Indian children 2) To assess which of the three methods of BA assessment is more suitable in Indian children.Methodology:X-rays of 851 children (438 boys and 413 girls, aged 2–16.5 years) were analyzed by four independent observers using three different methods of BA estimation (GP, GR, and TW-3). Mean BAs were converted to Z-scores. For purpose of deciding which method of BA was most suitable in our cohort, a test of proportions and root mean square (RMS) deviations were computed.Results:Using the test of proportions, the TW-3 method was most suitable overall (P < 0.05). TW-3 method was again most applicable in prepubertal boys (P < 0.05), in prepubertal girls (although not significant, P > 0.1), and pubertal girls (P < 0.05). However, in pubertal boys, the GR atlas method was most suitable (P < 0.05). The same results were obtained when root mean square (RMS) deviations were computed. Interestingly, BA was underestimated in Indian boys irrespective of the method used. In Indian girls, however, the BA was underestimated till the pubertal growth spurt, after which there was rapid advancement of BA.Conclusions:Among the three methods (GP, GR, and TW-3), the BAs estimated by the TW-3 method were closest to CAs. Hence, it seems reasonable to recommend the use of the TW-3 method for BA estimation in the Indian population till an Indian standard bone age atlas is developed.

Background: Accurate bone age assessment is essential for determining the actual degree of development and indicating a disorder in growth. While clinical bone age assessment techniques are time-consuming and prone to inter/intra-observer variability, deep learning-based methods are used for automated bone age estimation.Objective: The current study aimed to develop an unsupervised pre-training approach for automatic bone age estimation, addressing the challenge of limited labeled data and unique features of radiographic images of hand bones. Bone age estimation is complex and usually requires more labeling data. On the other hand, there is no model trained with hand radiographic images, reused for bone age estimationMaterial and Methods: In this fundamental-applied research, the collection of Radiological Society of North America (RSNA) X-ray image collection is used to evaluate the efficiency of the proposed bone age estimation method. An autoencoder is trained to reconstruct the original hand radiography images. Then, a model based on the trained encoder produces the final estimation of bone age.Results: Experimental results on the Radiological Society of North America (RSNA) X-ray image collection achieve a Mean Absolute Error (MAE) of 9.3 months, which is comparable to state-of-the-art methods. Conclusion: This study presents an approach to estimating bone age on hand radiographs utilizing unsupervised pre-training with an autoencoder and also highlights the significance of autoencoders and unsupervised learning as efficient substitutes for conventional techniques

Focus issue: Artificial intelligence in medical physics.

Bone age estimation is very useful in children undergoing epiphysiodesis or guided growth surgery especially during the years of accelerated growth. It may be noted that no data are available on bone age estimation for Indian children of this age group. Sauvegrain (French) method is a very useful and simple method for bone age assessment during the years of accelerated growth. We decided to check the usefulness and the accuracy of the Sauvegrain method in Indian children. A team of two pediatric orthopaedic surgeons and a radiologist scored elbow X-rays of 80 healthy children (40 boys and 40 girls), using the Sauvegrain method twice. Interobserver reliability and intraobserver reproducibility of the Sauvegrain scoring were assessed. There was a very strong correlation between all observers in both rounds (r = > 0.8) and an excellent reproducibility by the same observer in both rounds (r = 0.955). Chronological and bone age are considered the same if the difference between them is less than 6 months. With this criterion bone and chronological ages matched in >37% of boys and girls, similar to the study done in French children. In the nonmatching group, more children had delayed bone age compared to their chronological age. The Sauvegrain method of bone age assessment described for French children was found to be useful in estimating bone age in Indian children. It is especially helpful in the clinical practice for detecting mismatch between the chronological and the radiological age before undertaking guided growth or epiphysiodesis.

Bone age assessment is a common clinical procedure to diagnose endocrine and metabolic disorders in children. Recently, a variety of convolutional neural network based approaches have been developed to automatically estimate bone age from hand radiographs and achieved accuracy comparable to human experts. However, most of these networks were trained end-to-end, i.e., deriving the bone age directly from the whole input hand image without knowing which regions of the image are most relevant to the task. In this work, we proposed a multi-task convolutional neural network to simultaneously estimate bone age and localize ossification centers of different phalangeal, metacarpal and carpal bones. We showed that, similar to providing attention maps, the localization of ossification centers helps the network to extract features from more meaningful regions where local appearances are closely related to the skeletal maturity. In particular, to address the problem that some ossification centers do not always appear on the hand radiographs of certain bone ages, we introduced an image-level landmark presence classification loss, in addition to the conventional pixel-level landmark localization loss, in our multi-task network framework. Experiments on public RSNA data demonstrated the effectiveness of our proposed method in the reduction of gross errors of ossification center detection, as well as the improvement of bone age assessment accuracy with the aid of ossification center detection especially when the training data size is relatively small.

Aim and Objectives: For determining the skeletal/bone age from hand-wrist radiographs, Greulich and Pyle’s Atlas and Tanner–Whitehouse 3 (TW3) methods are considered as reliable methods, but there are concerns of extra radiation dose. The cervical vertebral bone age (CVBA) by measurements of the vertebral body of third (C3) and fourth (C4) is a relatively new method of objectively evaluating the skeletal maturation. The aim of this study was to estimate bone age by measuring 3rd and 4th cervical vertebral (C3, C4) dimensions. Materials and Methods: Lateral cephalometric and hand-wrist radiographs of 40 male and 46 female subjects (08–16 years of age) who had attended Government Dental College Srinagar for treatment were taken and measurements of C3 and C4 was done by using NNT Dicom version 7.0.0.0 software. Hand-wrist bone age was determined using Pyle and Greulich Atlas. Stepwise multiple regression analysis demonstrated the correlation between the measurements. Results: Anterior height of the third cervical vertebrae (AH3) in males had the strongest correlation with hand-wrist bone age (r=1.o, p3 was 10.733. The highest correlation coefficient in females was observed between H3 and the estimated bone age from Greulich’s Atlas (r=0.896, p3 was 10.43. The bone age was calculated by multiple regression model: Bone age in males = 0.809×AH3+ 2.896. Adjusted R2 was 0.153 (p3+ 2.752. Adjusted R2 was 0.245 (p Conclusions: Lateral cephalometric radiographs might be an alternative to the hand-wrist radiography for bone age estimation. Therefore, analysis of cervical vertebrae by replacing hand-wrist method additional radiograph could be avoided. Keywords: Cervical vertebral bone age, Lateral cephalometric radiographs, Hand-wrist radiographs.