Learning from imperfect training data using a robust loss function: application to brain image segmentation

TL;DR

This paper introduces a deep learning framework for brain MRI segmentation that is robust to noisy labels, improving accuracy in clinical applications like brain structure analysis and EEG/MEG source reconstruction.

Contribution

It presents a novel robust loss function enabling effective training with imperfect labels in brain image segmentation tasks.

Findings

The method achieves higher segmentation accuracy with noisy labels.

Robust training improves model reliability in clinical settings.

Application to brain MRI enhances neuroimaging analysis.

Abstract

Segmentation is one of the most important tasks in MRI medical image analysis and is often the first and the most critical step in many clinical applications. In brain MRI analysis, head segmentation is commonly used for measuring and visualizing the brain's anatomical structures and is also a necessary step for other applications such as current-source reconstruction in electroencephalography and magnetoencephalography (EEG/MEG). Here we propose a deep learning framework that can segment brain, skull, and extra-cranial tissue using only T1-weighted MRI as input. In addition, we describe a robust method for training the model in the presence of noisy labels.