Multi-Contrast MRI Super-Resolution via a Multi-Stage Integration Network

Abstract

Super-resolution (SR) plays a crucial role in improving the image quality of magnetic resonance imaging (MRI). MRI produces multicontrast images and can provide a clear display of soft tissues. However, current super-resolution methods only employ a single contrast, or use a simple multi-contrast fusion mechanism, ignoring the rich relations among different contrasts, which are valuable for improving SR. In this work, we propose a multi-stage integration network (i.e., MINet) for multi-contrast MRI SR, which explicitly models the dependencies between multi-contrast images at different stages to guide image SR. In particular, our MINet first learns a hierarchical feature representation from multiple convolutional stages for each of different-contrast image. Subsequently, we introduce a multi-stage integration module to mine the comprehensive relations between the representations of the multicontrast images. Specifically, the module matches each representation with all other features, which are integrated in terms of their similarities to obtain an enriched representation. Extensive experiments on fastMRI and real-world clinical datasets demonstrate that 1) our MINet outperforms state-of-the-art multi-contrast SR methods in terms of various metrics and 2) our multi-stage integration module is able to excavate complex interactions among multi-contrast features at different stages, leading to improved target-image quality.

Cross-Modality High-Frequency Transformer for MR Image Super-Resolution

“…A series of traditional algorithms [15,31] attempt to explore prior context information from T1-weighted MR images for super-resolving T2weighted or spectroscopy MR images. [9,14,39] further devise CNN models to exploit such inter-modality context information. However, directly fusing features of multiple modalities via convolutions with small kernels can not sufficiently leverage the intermodality dependencies.…”

“…Another important domain knowledge is that when processing LR T2WI data, the high-resolution T1-weighted images (T1WI) can be used to provide rich inter-modality context priors, as 1) the complementary morphological information [9] captured by the T1WI can help infer the structural content of the T2WI , and 2) the acquisition of HR T1WI costs much less scanning time 1 . An example is also provided to demonstrate the efficacy of such inter-modality context prior in Fig.…”

Super resolution reconstruction of CT images based on multi-scale attention mechanism

“…This paper summarizes the shortcomings of existing CT image super-resolution reconstruction as follows: (1) CT image has low contrast and insufficient high-frequency details. The above methods can't make full use of the global and local feature information of the image, resulting in the unclear detail texture of the reconstructed CT image [7,11]. (2) Medical images are an important basis for doctors to screen and diagnose, and the existing super-resolution reconstruction methods may introduce unnecessary noise, which can't ensure the pathological invariance of CT images, affecting doctors' accurate diagnosis [30,32,33,42].…”