Motivation: Precise diagnosis of endometrial cancer (EC) is essential to avoid unnecessary treatments and optimize patient outcomes. Current methods lack specificity in distinguishing benign from malignant lesions non-invasively. Goal(s): This study assesses time-dependent diffusion MRI's ability to differentiate EC from normal endometrial tissue, focusing on microstructural imaging biomarkers. Approach: A cohort of 22 patients underwent MRI using OGSE and PGSE sequences, and the diagnostic value of parameters like Vin and cellularity was evaluated. Results: Vin showed the highest diagnostic performance (AUC 0.867), followed by cellularity (AUC 0.829), demostrating their potential as reliable indicators of EC. Impact: Time-dependent diffusion MRI can non-invasively reveal microstructural characteristics of endometrial cancer, potentially improving diagnostic accuracy and informing preoperative planning, which may reduce invasive procedures and enhance treatment decision-making.
Yang et al. (Tue,) studied this question.