5 papers
When Oracle Conditioning Misleads Deployment: Conditioning-Availability Bias in Echocardiographic Segmentation
Dang P. M. Cao, Hieu D. Pham, Hieu Pham
Conditional segmentation models may be trained and evaluated with auxiliary signals cleaner than those available at deployment. We study this protocol-level manifestation of shortc…
When Measurement Conventions Masquerade as Calibration Gains in Cardiac Digital Twins
Dang P. M. Cao, Hieu Pham
Cardiac digital twins convert clinical images into physiological measurements through observation operators, yet calibration studies often assume a fixed reference convention. Acro…
Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
Giang Nguyen, Raghav Mehta, Emma A. M. Stanley +4
Foundation models are increasingly used as image feature extractors for mammography, but their robustness under external domain shift remains unclear. We benchmark 15 foundation-mo…
LightX3ECG: A Lightweight and eXplainable Deep Learning System for 3-lead Electrocardiogram Classification
Khiem H. Le, Hieu H. Pham, Thao BT. Nguyen +3
Cardiovascular diseases (CVDs) are a group of heart and blood vessel disorders that is one of the most serious dangers to human health, and the number of such patients is still gro…
VinDr-CXR-VQA: A Visual Question Answering Dataset for Explainable Chest X-Ray Analysis with Multi-Task Learning
Dang H. Nguyen, Hieu H. Pham, Hao T. Nguyen
We present VinDr-CXR-VQA, a large-scale chest X-ray dataset for explainable Medical Visual Question Answering (Med-VQA) with spatial grounding. The dataset contains 17,597 question…