Zero-Shot Lung Disease Detection Using Radiological Symptomatic Descriptors and Pretrained Neural Networks.
Authors
Affiliations (4)
Affiliations (4)
- Erik Jonsson School of Engineering and Computer Science, The University of Texas at Dallas, Richardson, TX, USA.
- Institute of Information Technology, Jahangirnagar University, Dhaka, 1342, Bangladesh.
- Department of Information Technology, Faculty of Computing & Information Technology, King Abdulaziz University, Jeddah, 21589, Kingdom of Saudi Arabia.
- Institute of Information Technology, Jahangirnagar University, Dhaka, 1342, Bangladesh. [email protected].
Abstract
Aligning radiological features with clinical text descriptions remains a key challenge for zero-shot disease recognition in chest radiography. We propose DVLM (Dual-Head Vision-Language Model with Neural Memory), a framework combining Vision Transformer visual encoding with ClinicalBERT-based text processing through parallel contrastive and supervised learning branches. A neural memory module stores disease-relevant patterns during training for improved generalization to unseen pathologies. We evaluated DVLM on CheXpert, MIMIC-CXR, and PadChest using multi-seed validation (five seeds <math xmlns="http://www.w3.org/1998/Math/MathML"><mo>Ć</mo></math> fivefold cross-validation), controlled ablation studies, and statistical significance testing. DVLM achieved 90.0% Ā± 0.28% macro-averaged AUROC on CheXpert (95% CI, 89.5-90.6%), with the neural memory module contributing +3.3% improvement ( <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mi>p</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> , Cohen's <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mi>d</mi> <mo>=</mo> <mn>0.89</mn></mrow> </math> ). For zero-shot classification (25% held-out diseases), DVLM achieved 73.5% AUROC, outperforming MedKLIP by 2.3%. Temperature scaling reduced calibration error by 72%, and Grad-CAM localization achieved an IoU of 0.642 against radiologist annotations. Subgroup analysis confirmed equitable performance across demographic groups (maximum disparity, 1.3%). While DVLM demonstrates strong ranking capability suitable for triage applications, threshold-based classification for rare diseases remains limited (F1, 24.8-30.1%), indicating the need for radiologist confirmation in clinical deployment.