Improving Quality Control Of MRI Images Using Synthetic Motion Data

TL;DR

This paper presents a novel approach for MRI quality control that uses synthetic motion artifacts for pretraining, significantly enhancing accuracy and efficiency in automated QC systems.

Contribution

The study introduces synthetic motion data pretraining and transfer learning to improve MRI QC accuracy and reduce training resources, addressing dataset limitations.

Findings

Improved accuracy in MRI QC classification.

Reduced training time and computational resources.

Enhanced robustness of QC models with synthetic data.

Abstract

MRI quality control (QC) is challenging due to unbalanced and limited datasets, as well as subjective scoring, which hinder the development of reliable automated QC systems. To address these issues, we introduce an approach that pretrains a model on synthetically generated motion artifacts before applying transfer learning for QC classification. This method not only improves the accuracy in identifying poor-quality scans but also reduces training time and resource requirements compared to training from scratch. By leveraging synthetic data, we provide a more robust and resource-efficient solution for QC automation in MRI, paving the way for broader adoption in diverse research settings.