A large-scale heterogeneous 3D magnetic resonance brain imaging dataset for self-supervised learning

TL;DR

FOMO260K is a comprehensive large-scale heterogeneous MRI dataset designed to facilitate self-supervised learning research in medical imaging, including diverse scans and minimal preprocessing.

Contribution

The paper introduces FOMO260K, a new extensive MRI dataset with accompanying code and pretrained models for self-supervised learning.

Findings

Dataset includes 260,927 scans from diverse sources.

Supports development of self-supervised learning methods.

Provides pretrained models and code for benchmarking.

Abstract

We present FOMO260K, a large-scale, heterogeneous dataset of 260,927 brain Magnetic Resonance Imaging (MRI) scans from 77,589 MRI sessions and 55,378 subjects, aggregated from 910 publicly available sources. The dataset includes both clinical- and research-grade images, multiple MRI sequences, and a wide range of anatomical and pathological variability, including scans with large brain anomalies. Minimal preprocessing was applied to preserve the original image characteristics while reducing entry barriers for new users. Companion code for self-supervised pretraining and finetuning is provided, along with pretrained models. FOMO260K is intended to support the development and benchmarking of self-supervised learning methods in medical imaging at scale.