Zero-shot Bias Correction: Efficient MR Image Inhomogeneity Reduction Without Any Data

TL;DR

This paper introduces a zero-shot deep learning approach for MRI bias correction that does not require pre-training data, offering an efficient and accurate alternative to existing data-free methods.

Contribution

The authors propose a lightweight CNN for zero-shot bias correction in MRI images, eliminating the need for training data and improving efficiency and accuracy.

Findings

Outperforms current data-free N4 methods in accuracy.

Offers faster bias correction without pre-training.

Ensures stable convergence through iterative refinement.

Abstract

In recent years, deep neural networks for image inhomogeneity reduction have shown promising results. However, current methods with (un)supervised solutions require preparing a training dataset, which is expensive and laborious for data collection. In this work, we demonstrate a novel zero-shot deep neural networks, which requires no data for pre-training and dedicated assumption of the bias field. The designed light-weight CNN enables an efficient zero-shot adaptation for bias-corrupted image correction. Our method provides a novel solution to mitigate the biased corrupted image as iterative homogeneity refinement, which therefore ensures the considered issue can be solved easier with stable convergence of zero-shot optimization. Extensive comparison on different datasets show that the proposed method performs better than current data-free N4 methods in both efficiency and accuracy.