Machine learning-based personalized prediction of developmental dysplasia of the hip using risk factors and clinical examination.
7 affiliations
- 1Department of Medical Ultrasound, Jinhua Municipal Central Hospital, Zhejiang, China.
- 2Department of Medicine, Surgery and Pharmacy, University of Sassari, Sassari, Italy.
- 3Orthopedic and Traumatology Unit, Azienda Ospedaliero-Universitaria di Sassari, Sassari, Italy.
- 4Department of Medical Ultrasound, The Central Hospital of Wuhan, Tongji Medical College, Huazhong University of Science and Technology, Hubei, China.
- 5Department of Medical Ultrasonics, Affiliated Shenzhen Children's Hospital of Shantou University Medical College, Shenzhen, China.
- 6Department of Orthopedics, Affiliated Shenzhen Children's Hospital of Shantou University Medical College, Shenzhen, China.
- 7Department of Medical Ultrasonics, Shenzhen Children's Hospital & Shenzhen University Medical School, Shenzhen, China.
Abstract
Machine learning-based disease prediction models can assist with disease diagnosis and early detection. This study aimed to identify the risk factors for developmental dysplasia of the hip (DDH) and develop multiple machine learning-based models to support screening and risk stratification for DDH. A retrospective analysis of 12,105 infants screened for DDH between January 2020 and December 2022 at Shenzhen Children's Hospital was performed to collect information on physical examination and risk factors. After univariate and multivariate screening, age-stratified sampling was performed on 11,898 DDH-negative patients to match the age distribution of 207 DDH-positive patients, thereby reducing potential bias. These balanced cohorts were randomly split into training and test sets (at an 8:2 ratio) under both 1:1 and 1:2 positive-to-negative ratios. Predictive modeling was based on random forest, support vector machine, logistic regression, k-nearest neighbor, light gradient boosting machine, and extreme gradient boosting (XGBoost). Twelve variables were included in the model. The XGBoost model trained on 1:1 data was chosen as the best model [area under the curve (AUC): 0.883, accuracy: 0.807, precision: 0.744, F1 score: 0.800, recall (sensitivity): 0.865, and specificity: 0.761]. External validation using data from Jinhua Municipal Central Hospital provided preliminary evidence of transferability [AUC: 0.909, accuracy: 0.838, precision: 0.829, F1 score: 0.840, recall (sensitivity): 0.850, and specificity: 0.825]. Skin pattern asymmetry, restriction of hip abduction, swaddling, breech presentation, and female sex were the top five most important predictive variables. The XGBoost model can be used to support early clinical screening and risk stratification for DDH, helping clinicians identify infants who should undergo hip ultrasonography.
Related
- Prediction of chemical miscarriage in assisted reproductive technology pregnancies using machine learning: A developmental study
- Comparison of Open and Closed Reduction for Developmental Dysplasia of the Hip: A Retrospective Evaluation of Acetabular Development and Complications
- Perinatal and gestational factors associated with Pavlik harness outcomes in developmental dysplasia of the hip: a retrospective cohort study
- The Sacro-Femoral-Pubic Angle May Estimate Pelvic Tilt in Patients with Severe Developmental Dysplasia of the Hip and Leg Length Discrepancy Undergoing Total Hip Arthroplasty
- DeepOSSA: a deep learning framework for personalized survival prediction in osteosarcoma patients
BibTeX
@article{zhang2026machine,
title = {{Machine learning-based personalized prediction of developmental dysplasia of the hip using risk factors and clinical examination.}},
author = {Shuang-Shuang Zhang and Gianfilippo Caggiari and Xiao-Yi Chen and Bing-Xuan Huang and Han-Sheng Deng and Na Xu and Wei Shi},
journal = {Frontiers in digital health},
year = {2026},
month = oct,
doi = {10.3389/fdgth.2026.1912106},
url = {https://europepmc.org/article/MED/42953024},
note = {CC BY},
}
RIS
TY - JOUR
TI - Machine learning-based personalized prediction of developmental dysplasia of the hip using risk factors and clinical examination.
AU - Shuang-Shuang Zhang
AU - Gianfilippo Caggiari
AU - Xiao-Yi Chen
AU - Bing-Xuan Huang
AU - Han-Sheng Deng
AU - Na Xu
AU - Wei Shi
PY - 2026
DA - 2026/10/14
JO - Frontiers in digital health
DO - 10.3389/fdgth.2026.1912106
UR - https://europepmc.org/article/MED/42953024
AB - Machine learning-based disease prediction models can assist with disease diagnosis and early detection. This study aimed to identify the risk factors for developmental dysplasia of the hip (DDH) and develop multiple machine learning-based models to support screening and risk stratification for DDH. A retrospective analysis of 12,105 infants screened for DDH between January 2020 and December 2022 at Shenzhen Children's Hospital was performed to collect information on physical examination and risk factors. After univariate and multivariate screening, age-stratified sampling was performed on 11,898 DDH-negative patients to match the age distribution of 207 DDH-positive patients, thereby reducing potential bias. These balanced cohorts were randomly split into training and test sets (at an 8:2 ratio) under both 1:1 and 1:2 positive-to-negative ratios. Predictive modeling was based on random forest, support vector machine, logistic regression, k-nearest neighbor, light gradient boosting machine, and extreme gradient boosting (XGBoost). Twelve variables were included in the model. The XGBoost model trained on 1:1 data was chosen as the best model [area under the curve (AUC): 0.883, accuracy: 0.807, precision: 0.744, F1 score: 0.800, recall (sensitivity): 0.865, and specificity: 0.761]. External validation using data from Jinhua Municipal Central Hospital provided preliminary evidence of transferability [AUC: 0.909, accuracy: 0.838, precision: 0.829, F1 score: 0.840, recall (sensitivity): 0.850, and specificity: 0.825]. Skin pattern asymmetry, restriction of hip abduction, swaddling, breech presentation, and female sex were the top five most important predictive variables. The XGBoost model can be used to support early clinical screening and risk stratification for DDH, helping clinicians identify infants who should undergo hip ultrasonography.
SN - 2673-253X
ER -