DINO-LG: Enhancing Vision Transformers with Label Guidance for Coronary Artery Calcium Detection

TL;DR

This paper introduces DINO-LG, a label-guided self-supervised learning method for vision transformers that improves coronary artery calcium detection and scoring from CT scans, addressing data scarcity and imbalance issues.

Contribution

DINO-LG is a novel extension of DINO that incorporates label guidance and targeted augmentation, significantly enhancing CAC detection and scoring accuracy in medical imaging.

Findings

Achieved 89% sensitivity and 90% specificity in CAC slice detection.

Reduced false negatives by 49% and false positives by 57%.

Attained 90% accuracy in CAC risk classification.

Abstract

Coronary artery disease (CAD), one of the leading causes of mortality worldwide, necessitates effective risk assessment strategies, with coronary artery calcium (CAC) scoring via computed tomography (CT) being a key method for prevention. Traditional methods, primarily based on UNET architectures implemented on pre-built models, face challenges like the scarcity of annotated CT scans containing CAC and imbalanced datasets, leading to reduced performance in segmentation and scoring tasks. In this study, we address these limitations by introducing DINO-LG, a novel label-guided extension of DINO (self-distillation with no labels) that incorporates targeted augmentation on annotated calcified regions during self-supervised pre-training. Our three-stage pipeline integrates Vision Transformer (ViT-Base/8) feature extraction via DINO-LG trained on 914 CT scans comprising 700 gated and 214 non-gated acquisitions, linear classification to identify calcified slices, and U-NET segmentation for CAC quantification and Agatston scoring. DINO-LG achieved 89% sensitivity and 90% specificity for detecting CAC-containing CT slices, compared to standard DINO's 79% sensitivity and 77% specificity, reducing false-negative and false-positive rates by 49% and 57% respectively. The integrated system achieves 90% accuracy in CAC risk classification on 45 test patients, outperforming standalone U-NET segmentation (76% accuracy) while processing only the relevant subset of CT slices. This targeted approach enhances CAC scoring accuracy by feeding the UNET model with relevant slices, improving diagnostic precision while lowering healthcare costs by minimizing unnecessary tests and treatments.