Multiple Instance Neuroimage Transformer

TL;DR

This paper introduces MINiT, a novel convolution-free transformer model utilizing multiple instance learning for classifying T1-weighted MRIs, demonstrating its effectiveness in identifying sex differences in brain structure.

Contribution

The paper presents the first application of multiple instance learning with transformers for neuroimage classification, specifically for T1-weighted MRI analysis.

Findings

MINiT effectively identifies sex differences in brain MRI.

Attention maps highlight relevant brain voxels for sex classification.

Model achieves promising results on public neuroimaging datasets.

Abstract

For the first time, we propose using a multiple instance learning based convolution-free transformer model, called Multiple Instance Neuroimage Transformer (MINiT), for the classification of T1weighted (T1w) MRIs. We first present several variants of transformer models adopted for neuroimages. These models extract non-overlapping 3D blocks from the input volume and perform multi-headed self-attention on a sequence of their linear projections. MINiT, on the other hand, treats each of the non-overlapping 3D blocks of the input MRI as its own instance, splitting it further into non-overlapping 3D patches, on which multi-headed self-attention is computed. As a proof-of-concept, we evaluate the efficacy of our model by training it to identify sex from T1w-MRIs of two public datasets: Adolescent Brain Cognitive Development (ABCD) and the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA). The learned attention maps highlight voxels contributing to identifying sex differences in brain morphometry. The code is available at https://github.com/singlaayush/MINIT.