What question did this study set out to answer?

The aim is to develop an automated method for assessing the quality of under-sampled MRI reconstructions without needing fully-sampled reference images.

July 28, 2026Magnetic Resonance in Medicine

Pixel‐Wise Uncertainty Quantification of Accelerated MRI Reconstruction

View Full Paper

Ask AI

Bookmark

Authors

IGIlias I. GiannakopoulosNew York University LMLokesh B. Gautham MuthukumarCourant Institute of Mathematical Sciences YLYvonne W. LuiUniversity of Maryland, Baltimore

Discussion

Member takes

Implication

Randomized trial evaluates pixel-wise uncertainty in MRI images, implying improved diagnostic reliability.

Key Points

The aim is to develop an automated method for assessing the quality of under-sampled MRI reconstructions without needing fully-sampled reference images.
Introduced a framework for pixel-wise uncertainty quantification in accelerated MRI.
Utilized conformal quantile regression with learning-based reconstruction methods for estimating uncertainties.
Trained and evaluated on MRI data from the fastMRI dataset with acceleration factors from 2 to 10.
Predicted uncertainty maps showed a Pearson correlation coefficient greater than 90% at four-fold and above acceleration levels.
Correlation dropped to below 70% when using a simpler heuristic for uncertainty computation.
Uncertainty maps effectively captured the magnitude and distribution of reconstruction errors, aligning with detected pathologies and artifacts.

Cite This Study

Giannakopoulos et al. (2026) studied this question.

synapsesocial.com/papers/6a6850fb2845b684d16f026c https://doi.org/10.1002/mrm.70526

View Full Paper

Ask AI

Bookmark

Pixel‐Wise Uncertainty Quantification of Accelerated MRI Reconstruction

Discussion

Member takes

Implication

Key Points

Cite This Study

Pixel‐Wise Uncertainty Quantification of Accelerated MRI Reconstruction

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider