A deep-learning model can segment and localize various devices on pediatric chest x-rays with promising accuracy.
Key Details
- 1Study evaluated 1,000 chest x-rays from 553 neonatal and pediatric patients (median age: 118 days).
- 2DeepLabV3+ architecture used to segment endotracheal tubes, tracheostomy tubes, and left/right PICC lines.
- 3Model achieved a mean absolute error (MAE) of 7.58 mm and mean Dice similarity coefficient (DSC) of 0.70 overall.
- 4Endotracheal tube localization was most accurate (MAE 5.79 mm, DSC 0.78).
- 5Manual annotation performed by image analysts and radiologists established ground truth.
- 6Automated approach could ease radiologist workload in intensive care settings.
Why It Matters
Efficient and accurate localization of devices on pediatric x-rays is critical in intensive care, where manual review is time-consuming. AI automation could reduce workload and potentially improve patient safety in these high-pressure environments.

Source
AuntMinnie
Related News

•AuntMinnie
Deep Learning PET Biomarkers Outperform Clinical Tools in Prostate Cancer Survival Prediction
Deep learning-derived PET biomarkers more accurately predict metastatic prostate cancer survival than standard clinical risk models.

•AuntMinnie
Multinational AI Model Flags Osteoporosis on Chest X-rays
AI can opportunistically detect osteoporosis from chest x-rays across diverse international populations.

•AuntMinnie
LLM-Based Chart Summarization Tool Reduces Radiologist Review Time
An LLM-based chart summarization tool safely reduced radiologist study review time by 17%.