GIIM: Graph-based Learning of Inter- and Intra-view Dependencies for Multi-view Medical Image Diagnosis

TL;DR

GIIM is a graph-based model that captures intra- and inter-view dependencies in multi-view medical images, improving diagnosis accuracy and robustness, especially with incomplete data.

Contribution

This paper introduces GIIM, a novel graph-based framework that models complex intra- and inter-view relationships and handles missing data in multi-view medical image diagnosis.

Findings

Significantly improves diagnostic accuracy over existing methods.

Effectively handles incomplete data in multi-view settings.

Validated across CT, MRI, and mammography datasets.

Abstract

Computer-aided diagnosis (CADx) has become vital in medical imaging, but automated systems often struggle to replicate the nuanced process of clinical interpretation. Expert diagnosis requires a comprehensive analysis of how abnormalities relate to each other across various views and time points, but current multi-view CADx methods frequently overlook these complex dependencies. Specifically, they fail to model the crucial relationships within a single view and the dynamic changes lesions exhibit across different views. This limitation, combined with the common challenge of incomplete data, greatly reduces their predictive reliability. To address these gaps, we reframe the diagnostic task as one of relationship modeling and propose GIIM, a novel graph-based approach. Our framework is uniquely designed to simultaneously capture both critical intra-view dependencies between abnormalities and inter-view dynamics. Furthermore, it ensures diagnostic robustness by incorporating specific techniques to effectively handle missing data, a common clinical issue. We demonstrate the generality of this approach through extensive evaluations on diverse imaging modalities, including CT, MRI, and mammography. The results confirm that our GIIM model significantly enhances diagnostic accuracy and robustness over existing methods, establishing a more effective framework for future CADx systems.