ASMFS: Adaptive-Similarity-based Multi-modality Feature Selection for Classification of Alzheimer's Disease

TL;DR

This paper introduces ASMFS, a novel multi-modality feature selection method that jointly learns similarity matrices across different imaging modalities and performs feature selection, improving Alzheimer's disease classification accuracy.

Contribution

The paper proposes a joint learning framework for multi-modality feature selection and similarity matrix learning, addressing limitations of fixed similarity matrices in prior methods.

Findings

Outperforms existing multi-modality approaches on ADNI dataset

Demonstrates improved classification accuracy for Alzheimer's disease

Validates effectiveness of joint similarity learning and feature selection

Abstract

With the increasing amounts of high-dimensional heterogeneous data to be processed, multi-modality feature selection has become an important research direction in medical image analysis. Traditional methods usually depict the data structure using fixed and predefined similarity matrix for each modality separately, without considering the potential relationship structure across different modalities. In this paper, we propose a novel multi-modality feature selection method, which performs feature selection and local similarity learning simultaniously. Specially, a similarity matrix is learned by jointly considering different imaging modalities. And at the same time, feature selection is conducted by imposing sparse l_{2, 1} norm constraint. The effectiveness of our proposed joint learning method can be well demonstrated by the experimental results on Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, which outperforms existing the state-of-the-art multi-modality approaches.