Accelerating Whole-Body Diffusion-weighted MRI with Deep Learning-based Denoising Image Filters

Purpose: To use deep learning to improve the image quality of subsampled images (number of acquisitions = 1 [NOA 1]) to reduce whole-body diffusion-weighted MRI (WBDWI) acquisition times. Materials and Methods: Both retrospective and prospective patient groups were used to develop a deep learning-based denoising image filter (DNIF) model. For initial model training and validation, 17 patients with metastatic prostate cancer with acquired WBDWI NOA 1 and NOA 9 images (acquisition period, 2015-2017) were retrospectively included. An additional 22 prospective patients with advanced prostate cancer, myeloma, and advanced breast cancer were used for model testing (2019), and the radiologic quality of DNIF-processed NOA 1 (NOA 1-DNIF) images were compared with NOA 1 images and clinical NOA 16 images by using a three-point Likert scale (good, average, or poor; statistical significance was calculated by using a Wilcoxon signed ranked test). The model was also retrained and tested in 28 patients with malignant pleural mesothelioma (MPM) who underwent lung MRI (2015-2017) to demonstrate feasibility in other body regions. Results: The model visually improved the quality of NOA 1 images in all test patients, with the majority of NOA 1-DNIF and NOA 16 images being graded as either average or good across all image-quality criteria. From validation data, the mean apparent diffusion coefficient (ADC) values within NOA 1-DNIF images of bone disease deviated from those within NOA 9 images by an average of 1.9% (range, 1.1%-2.6%). The model was also successfully applied in the context of MPM; the mean ADCs from NOA 1-DNIF images of MPM deviated from those measured by using clinical-standard images (NOA 12) by 3.7% (range, 0.2%-10.6%). Conclusion: Clinical-standard images were generated from subsampled images by using a DNIF. Supplemental material is available for this article. Published under a CC BY 4.0 license.

Automated summary

Utilizing deep learning, the study successfully improved the quality of subsampled images, with the majority of processed images achieving good quality ratings. The model's effectiveness was demonstrated across various cases, showing promising potential for clinical applications in reducing image acquisition times without compromising image quality.