On the Feasibility of Machine Learning Augmented Magnetic Resonance for Point-of-Care Identification of Disease

TL;DR

This paper investigates an ML-augmented MRI approach that directly infers disease presence from minimal k-space data, bypassing image reconstruction, thus enabling faster, more accessible point-of-care disease detection.

Contribution

It introduces a method to select optimal k-space subsets for disease classification, achieving comparable accuracy with significantly less data and no image reconstruction.

Findings

Achieves disease detection with only 8% of k-space data for prostate and brain scans.

Detects knee abnormalities using just 5% of the k-space data.

Provides an extensive analysis and open-source code for future research.

Abstract

Early detection of many life-threatening diseases (e.g., prostate and breast cancer) within at-risk population can improve clinical outcomes and reduce cost of care. While numerous disease-specific "screening" tests that are closer to Point-of-Care (POC) are in use for this task, their low specificity results in unnecessary biopsies, leading to avoidable patient trauma and wasteful healthcare spending. On the other hand, despite the high accuracy of Magnetic Resonance (MR) imaging in disease diagnosis, it is not used as a POC disease identification tool because of poor accessibility. The root cause of poor accessibility of MR stems from the requirement to reconstruct high-fidelity images, as it necessitates a lengthy and complex process of acquiring large quantities of high-quality k-space measurements. In this study we explore the feasibility of an ML-augmented MR pipeline that directly infers the disease sidestepping the image reconstruction process. We hypothesise that the disease classification task can be solved using a very small tailored subset of k-space data, compared to image reconstruction. Towards that end, we propose a method that performs two tasks: 1) identifies a subset of the k-space that maximizes disease identification accuracy, and 2) infers the disease directly using the identified k-space subset, bypassing the image reconstruction step. We validate our hypothesis by measuring the performance of the proposed system across multiple diseases and anatomies. We show that comparable performance to image-based classifiers, trained on images reconstructed with full k-space data, can be achieved using small quantities of data: 8% of the data for detecting multiple abnormalities in prostate and brain scans, and 5% of the data for knee abnormalities. To better understand the proposed approach and instigate future research, we provide an extensive analysis and release code.