# Evaluating the Safety of Deep Learning-Based Brain MRI Reconstruction

arXiv: https://arxiv.org/abs/2608.28714 · HTML: https://mlsanity.com/paper/2608.28714/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 51 of its 299 references that are in the map.
- First posted on arXiv 28 Aug 2026.

## Builds on (most cited first, 30 of 51)

- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [MoDL: Model Based Deep Learning Architecture for Inverse Problems](https://mlsanity.com/paper/1712.02862/) (2017)
- [Learning a Variational Network for Reconstruction of Accelerated MRI Data](https://mlsanity.com/paper/1704.00447/) (2017)
- [Diffusion Models for Medical Image Analysis: A Comprehensive Survey](https://mlsanity.com/paper/2211.07804/) (2022)
- [A Deep Cascade of Convolutional Neural Networks for Dynamic MR Image Reconstruction](https://mlsanity.com/paper/1704.02422/) (2017)
- [On instabilities of deep learning in image reconstruction - Does AI come at a cost?](https://mlsanity.com/paper/1902.05300/) (2019)
- [Causality matters in medical imaging](https://mlsanity.com/paper/1912.08142/) (2019)
- [Adaptive Diffusion Priors for Accelerated MRI Reconstruction](https://mlsanity.com/paper/2207.05876/) (2022)
- [ISLES 2022: A multi-center magnetic resonance imaging stroke lesion segmentation dataset](https://mlsanity.com/paper/2206.06694/) (2022)
- [Deep Learning Methods for Parallel Magnetic Resonance Image Reconstruction](https://mlsanity.com/paper/1904.01112/) (2019)
- [On hallucinations in tomographic image reconstruction](https://mlsanity.com/paper/2012.00646/) (2020)
- [What is Healthy? Generative Counterfactual Diffusion for Lesion Localization](https://mlsanity.com/paper/2207.12268/) (2022)
- [The troublesome kernel -- On hallucinations, no free lunches and the accuracy-stability trade-off in inverse problems](https://mlsanity.com/paper/2001.01258/) (2020)
- [Self-Supervised MRI Reconstruction with Unrolled Diffusion Models](https://mlsanity.com/paper/2306.16654/) (2023)
- [Guided Reconstruction with Conditioned Diffusion Models for Unsupervised Anomaly Detection in Brain MRIs](https://mlsanity.com/paper/2312.04215/) (2023)
- [Stacked U-Nets with Self-Assisted Priors Towards Robust Correction of Rigid Motion Artifact in Brain MRI](https://mlsanity.com/paper/2111.06401/) (2021)
- [Learning Fourier-Constrained Diffusion Bridges for MRI Reconstruction](https://mlsanity.com/paper/2308.01096/) (2023)
- [SADIR: Shape-Aware Diffusion Models for 3D Image Reconstruction](https://mlsanity.com/paper/2309.03335/) (2023)
- [Machine Learning in Magnetic Resonance Imaging: Image Reconstruction](https://mlsanity.com/paper/2012.05303/) (2020)
- [Diff-Def: Diffusion-Generated Deformation Fields for Conditional Atlases](https://mlsanity.com/paper/2403.16776/) (2024)
- [Regression is all you need for medical image translation](https://mlsanity.com/paper/2505.02048/) (2025)
- [Latent Causal Modeling for 3D Brain MRI Counterfactuals](https://mlsanity.com/paper/2409.05585/) (2024)
- [Alternating Learning Approach for Variational Networks and Undersampling Pattern in Parallel MRI Applications](https://mlsanity.com/paper/2110.14703/) (2021)
- [Sinogram upsampling using Primal-Dual UNet for undersampled CT and radial MRI reconstruction](https://mlsanity.com/paper/2112.13443/) (2021)
- [Distortion-Corrected Image Reconstruction with Deep Learning on an MRI-Linac](https://mlsanity.com/paper/2205.10993/) (2022)
- [Zero-DeepSub: Zero-Shot Deep Subspace Reconstruction for Rapid Multiparametric Quantitative MRI Using 3D-QALAS](https://mlsanity.com/paper/2307.01410/) (2023)
- [Model-based Synthetic Data-driven Learning (MOST-DL): Application in Single-shot T2 Mapping with Severe Head Motion Using Overlapping-echo Acquisition](https://mlsanity.com/paper/2107.14521/) (2021)
- [SISMIK for brain MRI: Deep-learning-based motion estimation and model-based motion correction in k-space](https://mlsanity.com/paper/2312.13220/) (2023)
- [NPB-REC: A Non-parametric Bayesian Deep-learning Approach for Undersampled MRI Reconstruction with Uncertainty Estimation](https://mlsanity.com/paper/2404.04550/) (2024)
- [Non-Reference Quality Assessment for Medical Imaging: Application to Synthetic Brain MRIs](https://mlsanity.com/paper/2407.14994/) (2024)

## Abstract (arXiv)

Objective: Deep learning accelerates brain MRI four- to tenfold, but models can erase lesions or synthesize false tissue - failures pixel-averaged metrics like PSNR and SSIM miss. We review whether current evaluation practices detect this blind spot. Methods: Following PRISMA 2020, we searched seven databases without date limits, including 263 studies (1995-2026), appraised them using QUADAS-2 and matched instruments, and synthesized narratively. Categories were derived from titles, abstracts, and controlled vocabulary; reported prevalence figures represent floors. Duplicate screening achieved high agreement (Fleiss kappa = 0.877), as did appraisal (0.788; 0.390 where observable). Extraction is unaudited. Results: Only 18 of 263 studies (6.8%) recorded both a fidelity metric and reader assessment on identical data, leaving the central surrogate unmeasured. Reader studies mostly measured inter-reader agreement, which was weak: fastMRI 2020 concordance reached 0.457 and 0.386 (Kendall W), improving only where SSIM diverged. Erasing a 100 mm3 lacunar infarct shifts global PSNR by 0.03 dB under the stated error model. As the corpus grew fivefold, reader assessments dropped from 32% to 18%, recovering to 21%. Generative models - most associated with hallucination (39%) - were among the least reader-evaluated (11.3%), while self-supervised models reached 47% with zero reader evaluation. Only 5% released code and ran reader studies; none evaluated a model observer; no named dataset covered acute stroke or hemorrhage. Conclusions: On these floors, current evaluation practices cannot certify diagnostic safety. We derive five requirements safety-oriented evaluations must meet.