What question did this study set out to answer?

This study aims to understand why deep learning models for quantitative MRI fail when trained on one vendor and used on another.

June 12, 2026Open Access

Why Deep Learning Fails Across MRI Vendors: A Physics Attribution Study with Dose-Response Curves, Scaling Laws, and Hybrid Rescues

Key Points

This study aims to understand why deep learning models for quantitative MRI fail when trained on one vendor and used on another.
Isolated 9 physical corruptions impacting deep learning-based parameter estimation from a vendor shift.
Generated 100,000 synthetic MRI signals using Bloch-equation simulation.
Tested 4 robustness algorithms across 2 architectures with validation on real multi-scanner MRF data from 30 brain scans.
B₀ inhomogeneity identified as the dominant corruption with a 25 Hz peak effect.
Without peak normalization, halving B₁ increases error by 39%.
Standard robustness algorithms failed significantly, while a hybrid approach outperformed both classical and deep learning techniques.

Abstract

When a deep learning model for quantitative MRI (qMRI) is trained on one scanner vendor and deployed on another, it often fails dramatically. But why does it fail? We build a framework to isolate each physical factor and measure exactly how much damage it causes to deep learning-based parameter estimation. We generate 100,000 synthetic MRF (Magnetic Resonance Fingerprinting) signals using Bloch-equation simulation, decompose vendor shift into 9 isolated physical corruptions, and test 4 robustness algorithms (ERM, DeepCORAL, GroupDRO, IRM) across 2 architectures (ResNet-1D, ViT-1D). We validate on real multi-scanner MRF data (30 brain scans across 3 scanners). Key findings:1. B₀ inhomogeneity is the dominant corruption with a non-monotonic dose-response peaking at 25 Hz.2. Peak normalization masks B₁⁺ sensitivity — without it, halving B₁ causes 39% error increase.3. All standard robustness algorithms fail (DS3 = 39–58×); IRM fails entirely.4. Data scaling cannot overcome uncalibrated physics — more source data does not improve out-of-distribution performance.5. A simple hybrid combining deep learning with classical dictionary matching outperforms either approach alone.6. Real multi-scanner data confirms T₂ is 3.5× more variable than T₁ across scanners. This repository contains the full paper (LaTeX source, 16 pages, 10 figures), source code/Github link(Bloch simulator, ResNet-1D, ViT-1D, ERM/CORAL/GroupDRO/IRM), 26 measured experiments (all checkpointed and resumable), and all result data in JSON format.

Read Full Paperexternally

Ask AI

Mark Helpful

Bookmark

Relay

View Full Paper