Improved Multimodal Fusion for Small Datasets with Auxiliary Supervision

TL;DR

This paper introduces three simple auxiliary supervision techniques to improve multimodal fusion in small datasets, demonstrated on prostate cancer diagnosis with image and clinical data, reducing overfitting and enhancing model performance.

Contribution

Proposes three auxiliary supervision methods for better multimodal fusion in small datasets, avoiding complex fusion operations and improving training stability.

Findings

Enhanced fusion methods outperform traditional concatenation.

Auxiliary supervision improves model generalization.

Applicable to various paired image and non-image classification tasks.

Abstract

Prostate cancer is one of the leading causes of cancer-related death in men worldwide. Like many cancers, diagnosis involves expert integration of heterogeneous patient information such as imaging, clinical risk factors, and more. For this reason, there have been many recent efforts toward deep multimodal fusion of image and non-image data for clinical decision tasks. Many of these studies propose methods to fuse learned features from each patient modality, providing significant downstream improvements with techniques like cross-modal attention gating, Kronecker product fusion, orthogonality regularization, and more. While these enhanced fusion operations can improve upon feature concatenation, they often come with an extremely high learning capacity, meaning they are likely to overfit when applied even to small or low-dimensional datasets. Rather than designing a highly expressive fusion operation, we propose three simple methods for improved multimodal fusion with small datasets that aid optimization by generating auxiliary sources of supervision during training: extra supervision, clinical prediction, and dense fusion. We validate the proposed approaches on prostate cancer diagnosis from paired histopathology imaging and tabular clinical features. The proposed methods are straightforward to implement and can be applied to any classification task with paired image and non-image data.