DA-VSR: Domain Adaptable Volumetric Super-Resolution for Medical Images

Abstract

Medical image super-resolution (SR) is an active research area that has many potential applications, including reducing scan time, bettering visual understanding, increasing robustness in downstream tasks, etc. However, applying deep-learning-based SR approaches for clinical applications often encounters issues of domain inconsistency, as the test data may be acquired by different machines or on different organs. In this work, we present a novel algorithm called domain adaptable volumetric super-resolution (DA-VSR) to better bridge the domain inconsistency gap. DA-VSR uses a unified feature extraction backbone and a series of network heads to improve image quality over different planes. Furthermore, DA-VSR leverages the in-plane and through-plane resolution differences on the test data to achieve a self-learned domain adaptation. As such, DA-VSR combines the advantages of a strong feature generator learned through supervised training and the ability to tune to the idiosyncrasies of the test volumes through unsupervised learning. Through experiments, we demonstrate that DA-VSR significantly improves superresolution quality across numerous datasets of different domains, thereby taking a further step toward real clinical applications.

RPLHR-CT Dataset and Transformer Baseline for Volumetric Super-Resolution from CT Scans

“…Fig. 3(a) summarizes the quantitative comparisons of our method and other state-of-the-art CT volumetric SR methods: ResVox [6], MPU-Net [13], SAINT [17] and DA-VSR [18]. For ResVox, the noise reduction part is removed.…”

“…CNN-based algorithms have achieved outstanding performance in SR for natural images [20] and these techniques have been introduced for volumetric SR [1,4,6,12,13,15,17,18,23,25]. Though significant advances have been made, CNN-based algorithms remain limited by the inherent weaknesses of convolution operators.…”

“…Another impediment to the application of volumetric SR methods is data. Most relevant studies use HR volume as ground truth and degrade it to construct paired pseudo-LR volumes with which to train and evaluate methods [4,15,17,18,23,25]. For instance, Peng et al [17] perform sparse sampling on the depth dimension of thin CT to obtain pseudo thick CT. Zhao et al [25] simulate pseudo-LR MRI by applying an ideal low-pass filter to the isotropic T2-weighted MRI followed by an anti-ringing Fermi filter.…”

Rethinking Degradation: Radiograph Super-Resolution via AID-SRGAN

“…High-resolution musculoskeletal radiographs provide more details that are crucial for medical diagnosis, particularly for diagnosing primary bone tumors and bone stress injuries [2,4,8,18,22]. However, radiographic image quality is affected by many factors, such as scanning time, patients' poses, and motions, and achieving higher-resolution medical images is expensive and time-consuming because it requires a relatively long scanning time.…”

“…To address this problem, more attention has been paid to the tasks of medical image super-resolution [7,18,25]. Deep learning-based methods [12,16,29,33,34] dominate image SR, which learns a mapping from LR images to HR images and differs from traditional methods in which more prior knowledge is required [3].…”