Better diagnostic value and feasibility of Deep Learning DWI in uterine malignant neoplasms

Motivation: Conventional diffusion-weighted imaging (c-DWI) of the uterus is time-consuming, and the lesion details are not well-defined. Goal(s): To introduce a deep learning (DL) DWI sequence in uterine MRI and compare it with conventional DWI (c-DWI) to investigate its impact on examination time, image quality, lesion significance, diagnostic reliability, as well as contrast ratio (CN), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). Approach: 10 patients with uterine malignancy disease were included in this study. Results: There is no significant difference in objective assessment between the two techniques, while the overall image quality of DL-DWI is better than c-DWI (p < 0.01). Impact: The research investigated the utilization of DL-DW in the uterus, which led to shorter examination times and significantly improved image quality. This analysis has the potential to examine other pelvic organs, such as the prostate, to assess pelvic lesions.