Adult-trained radiology AI models often underperform when applied to pediatric imaging data, according to a systematic review.
Key Details
- 1Researchers reviewed over 2,000 articles but only 15 met inclusion criteria for full-text analysis.
- 2Studied AI tasks included segmentation, object detection, and classification.
- 3Only 2 adult-trained models (organ segmentation on CT) showed similar accuracy between adults and children.
- 4Most models performed worse on pediatric data, with variable degrees of performance drop.
- 5Fine-tuning with limited pediatric data led to improvements in 5 out of 6 studies, but none outperformed the original adult-trained models.
- 6Findings highlight the need for dedicated pediatric imaging datasets for AI model development.
Why It Matters
Reliable AI model performance for pediatric imaging cannot be assumed based on adult-trained models alone. This underscores a critical demand for collecting pediatric-specific data to develop safe and effective AI tools for children.
Source
AuntMinnie
Related News
•Radiology Business
Meta Funds Ohio State’s Radiology AI Innovation Projects
Meta will fund Ohio State researchers to explore AI-driven advancements in radiology, including chest X-ray reporting and disease prediction.
•AuntMinnie
ChatGPT-5 Performs Poorly at Detecting Distal Radius Fractures on X-rays
ChatGPT-5 missed about half of distal radius fractures on x-rays, showing limited diagnostic capability in a real-world scenario.
•AuntMinnie
AI Triage Tool Speeds Up ICH Head CT Turnaround in ED
AI-based triage for intracranial hemorrhage on head CT reduced emergency department report turnaround by 13 minutes.