MMV-Net: A Multiple Measurement Vector Network for Multi-frequency Electrical Impedance Tomography

TL;DR

This paper introduces MMV-Net, a deep learning model that effectively reconstructs multi-frequency electrical impedance images by capturing frequency correlations, outperforming existing methods in quality, convergence, and noise robustness.

Contribution

The paper proposes MMV-Net, a novel deep learning framework that models multi-frequency EIT reconstruction using a MMV approach with attention and memory modules.

Findings

Superior image quality compared to state-of-the-art methods

Faster convergence and better noise robustness

Effective modeling of intra- and inter-frequency dependencies

Abstract

Multi-frequency Electrical Impedance Tomography (mfEIT) is an emerging biomedical imaging modality to reveal frequency-dependent conductivity distributions in biomedical applications. Conventional model-based image reconstruction methods suffer from low spatial resolution, unconstrained frequency correlation and high computational cost. Deep learning has been extensively applied in solving the EIT inverse problem in biomedical and industrial process imaging. However, most existing learning-based approaches deal with the single-frequency setup, which is inefficient and ineffective when extended to address the multi-frequency setup. In this paper, we present a Multiple Measurement Vector (MMV) model based learning algorithm named MMV-Net to solve the mfEIT image reconstruction problem. MMV-Net takes into account the correlations between mfEIT images and unfolds the update steps of the Alternating Direction Method of Multipliers (ADMM) for the MMV problem. The non-linear shrinkage operator associated with the weighted l2_1 regularization term is generalized with a cascade of a Spatial Self-Attention module and a Convolutional Long Short-Term Memory (ConvLSTM) module to capture intra- and inter-frequency dependencies. The proposed MMVNet was validated on our Edinburgh mfEIT Dataset and a series of comprehensive experiments. All reconstructed results show superior image quality, convergence performance and noise robustness against the state of the art.