Get Started
Home
Topics
Search
Library
Research questionHow can pediatric chest X-ray classifiers preserve usable decisions across countries when calibration and source thresholds shift?A classifier may retain some ability to rank pediatric chest X-rays from another country while its predicted probabilities and source-selected decision threshold become unreliable. Limited local adaptation can recover sensitivity without necessarily restoring specificity or a stable alert burden.
AI
Computer Vision
Evaluation & Benchmarks
Health
Image & Video Processing
Machine Learning
Research Paper
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.Cross-dataset transportability of pediatric chest X-ray deep learning across three countries: discrimination, calibration, operating-point failure, and limited-label recoveryThe evidence concerns a frozen three-seed DenseNet121 dual-view ensemble developed using 5,824 Guangzhou radiographs and evaluated zero-shot on cohorts from Bangladesh and Vietnam. Secondary analyses used 163–651 labeled Bangladesh images for adaptation and assessed discrimination, calibration, fixed-threshold behavior, shortcut-associated signal, and alert burden; conclusions are limited to pediatric pneumonia classification and these datasets and protocols.research paper · Sep 4, 2026
Related questions
Can multimodal chest-radiograph triage trained on NLP-derived labels reliably match expert severity judgments?How should federated aggregation weight hospitals when local data size does not reflect diagnostic informativeness?How does parameter-efficient adapter choice affect pathology accuracy and subgroup disparities when adapting a frozen chest X-ray encoder?How can early-stage chronic kidney disease screening remain accurate and stable with limited labeled data?