What type of study is this?

September 10, 2025Open Access

Artificial Intelligence Algorithm Supporting the Diagnosis of Developmental Dysplasia of the Hip: Automated Ultrasound Image Segmentation

Puntos clave

The machine learning model achieved high segmentation accuracy for hip structures, enhancing ultrasound diagnostics.
Model performance metrics like Intersection over Union and Dice Similarity indicate its effectiveness in identifying vital hip anatomy.
Utilizing a dataset of over 10,000 images, the model demonstrated superior performance with precise measurements for key hip structures.
Integration of real-time features may further support clinical decision-making in diagnosing developmental dysplasia of the hip.

Resumen

Background: Developmental dysplasia of the hip (DDH), if not treated, can lead to os-teoarthritis and disability. Ultrasound (US) is a primary screening method for the de-tection of DDH, but its interpretation remains highly operator-dependent. We propose a supervised machine learning (ML) image segmentation model for the automated recognition of anatomical structures in hip US images. Methods: A dataset of 10, 767 images from 311 patients was annotated for eight key structures according to the Graf method and split into training (75. 0%), validation (9. 5%), and test (15. 5%) sets. Model performance was assessed using the Intersection over Union (IoU) and Dice Similarity Coefficient (DSC). Results: The best-performing model was based on the SegNeXt ar-chitecture with an MSCANL backbone. The model achieved high segmentation accu-racy, (IoU; DSC) for chondro-osseous border (0. 632; 0. 774), femoral head (0. 916; 0. 956), labrum (0. 625; 0. 769), cartilaginous (0. 672; 0. 804) and bony roof (0. 725; 0. 841). The av-erage Euclidean distance for point-based landmarks (bony rim and lower limb) was 4. 8 and 4. 5 pixels, respectively, and the baseline deflection angle was 1. 7 degrees. Conclu-sions: This ML-based approach demonstrates promising accuracy and may enhance the reliability and accessibility of US-based DDH screening. Future applications could in-tegrate real-time angle measurement and automated classification to support clinical decision-making.

Leer artículo completoexternamente

Me gusta

Guardar

Ver artículo completo