MIST GAN: Modality Imputation Using Style Transfer for MRI

TL;DR

This paper introduces MIST GAN, a style transfer-based generative model that imputes missing MRI modalities from existing ones, aiming to reduce costs and improve clinical utility.

Contribution

It formulates MRI modality imputation as a style transfer problem using a multiple-to-one mapping network, demonstrating competitive results on BraTS'18 dataset.

Findings

Achieves results comparable to state-of-the-art in SSIM and PSNR

Model is validated by expert radiologists for clinical relevance

Analyzes style diversity within and across MRI modalities

Abstract

MRI entails a great amount of cost, time and effort for the generation of all the modalities that are recommended for efficient diagnosis and treatment planning. Recent advancements in deep learning research show that generative models have achieved substantial improvement in the aspects of style transfer and image synthesis. In this work, we formulate generating the missing MR modality from existing MR modalities as an imputation problem using style transfer. With a multiple-to-one mapping, we model a network that accommodates domain specific styles in generating the target image. We analyse the style diversity both within and across MR modalities. Our model is tested on the BraTS'18 dataset and the results obtained are observed to be on par with the state-of-the-art in terms of visual metrics, SSIM and PSNR. After being evaluated by two expert radiologists, we show that our model is efficient, extendable, and suitable for clinical applications.