Popular Boards
Suketu Bhavsar, Bhanu B. Gowda, Maulini Bhavsar | Journal of Paediatrics and Child Health | (2025)
Key Takeaways
Plain English Takeaway
Computer programs can help doctors spot hip problems in children just as well as experts can, but more testing in real-world settings is needed before using them everywhere.
Study Aim
The main goal of this paper is to systematically review how well artificial intelligence (AI) can detect developmental dysplasia of the hip (DDH), a condition where a child's hip joint does not form properly. The authors want to see if AI can match the accuracy of expert doctors when analyzing medical images like X-rays and ultrasounds for DDH diagnosis.
Simply put: The study checks if computer programs can find hip problems in kids as well as doctors do.
Study Design
The authors conducted a systematic review, following established guidelines for diagnostic test accuracy studies. They searched several medical and technical databases for studies up to October 2024. The review included 23 studies that used AI to analyze X-ray or ultrasound images of children under 16 years old for DDH. Only studies comparing AI results to expert doctors' diagnoses were included. The authors did not combine results into a meta-analysis because the studies used different methods and tools.
Simply put: The researchers looked at many studies where computers and doctors both checked kids' hip images, and compared their results.
Findings
The review found that AI systems can detect DDH on both X-ray and ultrasound images with high accuracy, sensitivity (ability to find true cases), and specificity (ability to rule out non-cases). In many studies, AI matched or even outperformed less experienced doctors and worked much faster than humans. However, the authors note that most studies used high-quality images and controlled settings, so results might not be as strong in real-world clinics. There was also a lot of variation in the types of AI used and the ages of children studied. The authors recommend more research using diverse, real-world data to make sure AI tools work well for all patients and settings.
Simply put: Computer programs can spot hip problems in kids very well in studies, but we need more real-life testing to be sure they work everywhere.
Abstract
AI demonstrated comparable effectiveness to physicians in detecting DDH. However, limited evaluation on external datasets restricts its generalisability. Further research incorporating diverse datasets and real-world applications is needed to assess its broader clinical impact on DDH diagnosis.
Referenced In
Mercedes C.
2 months ago
Created: Apr 11, 2026