Improving Deep Learning for Accelerated MRI With Data Filtering

TL;DR

This paper explores data filtering strategies to improve deep learning-based accelerated MRI reconstruction, demonstrating that curated training data enhances model robustness and performance across diverse datasets.

Contribution

It introduces data filtering methods for training deep neural networks on MRI data, showing consistent performance improvements across various conditions.

Findings

Filtering training data yields modest performance gains.

Filtering is especially beneficial with low in-distribution data.

Performance gains are consistent across different training set sizes and accelerations.

Abstract

Deep neural networks achieve state-of-the-art results for accelerated MRI reconstruction. Most research on deep learning based imaging focuses on improving neural network architectures trained and evaluated on fixed and homogeneous training and evaluation data. In this work, we investigate data curation strategies for improving MRI reconstruction. We assemble a large dataset of raw k-space data from 18 public sources consisting of 1.1M images and construct a diverse evaluation set comprising 48 test sets, capturing variations in anatomy, contrast, number of coils, and other key factors. We propose and study different data filtering strategies to enhance performance of current state-of-the-art neural networks for accelerated MRI reconstruction. Our experiments show that filtering the training data leads to consistent, albeit modest, performance gains. These performance gains are robust across different training set sizes and accelerations, and we find that filtering is particularly beneficial when the proportion of in-distribution data in the unfiltered training set is low.