Joint CS-MRI Reconstruction and Segmentation with a Unified Deep Network

TL;DR

This paper introduces SegNetMRI, a unified deep neural network that simultaneously reconstructs MRI images from compressed measurements and segments them, enhancing both tasks' performance.

Contribution

The paper presents a novel integrated deep learning architecture for joint CS-MRI reconstruction and segmentation, leveraging shared encoders for improved results.

Findings

SegNetMRI improves reconstruction quality with fewer measurements.

SegNetMRI enhances segmentation accuracy in compressed sensing MRI.

Joint training benefits both reconstruction and segmentation tasks.

Abstract

The need for fast acquisition and automatic analysis of MRI data is growing in the age of big data. Although compressed sensing magnetic resonance imaging (CS-MRI) has been studied to accelerate MRI by reducing k-space measurements, in current CS-MRI techniques MRI applications such as segmentation are overlooked when doing image reconstruction. In this paper, we test the utility of CS-MRI methods in automatic segmentation models and propose a unified deep neural network architecture called SegNetMRI which we apply to the combined CS-MRI reconstruction and segmentation problem. SegNetMRI is built upon a MRI reconstruction network with multiple cascaded blocks each containing an encoder-decoder unit and a data fidelity unit, and MRI segmentation networks having the same encoder-decoder structure. The two subnetworks are pre-trained and fine-tuned with shared reconstruction encoders. The outputs are merged into the final segmentation. Our experiments show that SegNetMRI can improve both the reconstruction and segmentation performance when using compressive measurements.