Significance of Normalized Apparent Diffusion Coefficient in the Vesical Imaging‐Reporting and Data System for Diagnosing Muscle‐Invasive Bladder Cancer

Background Vesical Imaging‐Reporting and Data System (VI‐RADS) has been developed for assessing bladder cancer from multiparametric (mp) MRI but its performance in diagnosing muscle‐invasive bladder cancer (MIBC) is suboptimal. Purpose To investigate associations between normalized apparent diffusion coefficient (NADC) and clinicopathological characteristics and to determine whether the inclusion of NADC can improve the performance of VI‐RADS in diagnosing MIBC. Study Type Retrospective. Population Two hundred seventy‐five patients with pathologically confirmed bladder cancer (101 MIBC and 174 non‐MIBC [NMIBC]) underwent preoperative mpMRI (233 male, 42 female). Field Strength/Sequence 3‐T, T2‐weighted imaging (turbo spin‐echo), diffusion‐weighted imaging (free‐breathing spin‐echo), and dynamic contrast‐enhanced imaging (gradient‐echo). Assessment NADC was the mean ADC of tumor divided by that of the iliopsoas muscles in trans caput femoris plane. Associations between NADC and clinicopathological characteristics were evaluated. Models were established for differentiating MIBC and NMIBC: VI‐RADS model; VN model (VI‐RADS and NADC), Images model (significant variables from imaging associated with MIBC), LN model (Images model without NADC), and Full model (all significant variables associated with MIBC). Statistical Tests Variables for model development were based on logistic regression. Models were evaluated by receiver operating characteristic (ROC) curve. Comparison of the area under the curves (AUCs) for the models used DeLong's test. A P value