Realistic noise synthesis reduces bias and improves tissue microstructure estimation with supervised machine learning
2026-06-01 • Machine Learning
Machine Learning
AI summaryⓘ
The authors studied how noise in MRI data can mess up machine learning models used to understand tiny tissue structures. They found that if the models are trained on simulated data without realistic noise, their predictions become biased, especially when the real data is noisy. To fix this, the authors created a new way to add realistic noise to training data, which made the models more accurate and less biased. This method worked well across different models and helps ensure better results, especially when scanning tissues with low signal quality.
Diffusion MRIMicrostructure parameter estimationSupervised machine learningNoise modelingRician noiseCovariate shiftSignal-to-noise ratio (SNR)MP-PCASpherical harmonicsNonlinear least-squares fitting
Authors
Bradley G. Karat, Maëliss Jallais, Ali R. Khan, Santiago Aja-Fernández, Jelle Veraart, Marco Palombo
Abstract
Diffusion MRI enables non-invasive probing of tissue microstructure, but accurate parameter estimation is challenged by noise-related effects. In supervised machine learning frameworks trained on simulated data, discrepancies between the noise characteristics of simulated and acquired signals introduce a form of covariate shift, whereby the input signal distribution differs between training and inference. We investigated the impact of this mismatch on microstructure parameter estimation and propose a realistic noise synthesis (RNS) framework to mitigate it. RNS incorporates both the Rician expectation and the effective post-processing noise variance into simulated training signals. The Rician expectation was modelled using a noise standard deviation estimated with MPPCA, while the effective standard deviation was derived from spherical harmonic residuals of preprocessed data. The method was evaluated using the cylinder-zeppelin and the SANDI models on simulated datasets across multiple SNR levels and on in vivo diffusion data with repeated acquisitions. Sensitivity to noise misestimation was also assessed. Ignoring magnitude-induced noise effects during training produced systematic, SNR-dependent parameter bias, particularly at low SNR. Incorporating the Rician expectation substantially reduced bias to the level of noise-aware nonlinear least-squares fitting. Modelling the effective standard deviation further improved precision. Performance was largely independent of regression architecture but sensitive to accurate noise estimation. These findings demonstrate that realistic noise modelling in simulated training data mitigates signal-domain covariate shift and is essential for unbiased supervised microstructure estimation, particularly in low-SNR regimes associated with high b-values or high spatial resolution.