Prospective Validation of Vesical Imaging-Reporting and Data System (VI-RADS) Using a Next-Generation Magnetic Resonance Imaging Scanner: Is Denoising Deep Learning Reconstruction Useful?

Researchers

Journal

Modalities

Models

Abstract

The Vesical Imaging Reporting and Data System (VI-RADS) was launched in 2018 to standardize reporting of magnetic resonance imaging (MRI) for bladder cancer (BC). This study aimed to prospectively validate VI-RADS using a next-generation MRI scanner and to investigate the usefulness of denoising deep learning reconstruction (dDLR).
We prospectively enrolled 98 patients who underwent bladder multiparametric MRI using a next-generation MRI scanner before transurethral resection of bladder tumor (TURBT). Tumors were categorized according to VI-RADS, and we ultimately analyzed 68 patients with pathologically confirmed urothelial BC. We used receiving operating characteristic curve analyses to assess the predictive accuracy of VI-RADS for muscle invasion. Sensitivity, specificity, positive/negative predictive value, accuracy, and area under the curve (AUC) were calculated for different VI-RADS score cutoffs.
Muscle invasion was detected in the TURBT specimens of 18 patients (26%). The optimal cutoff value of the VI-RADS score was determined as≥4 based on the receiver operating curve analyses. The accuracy of diagnosing muscle invasion using a cutoff of VI-RADS ≥4 was 94% (AUC: 0.92). Additionally, we assessed the utility of dDLR: combination with dDLR significantly improved the AUC of category by T2-weighted imaging (T2WI), and of the four patients who were misdiagnosed by the final VI-RADS score, three were correctly diagnosed by T2WI+dDLR.
In this prospective validation study with a next-generation MRI scanner, VI-RADS showed high predictive accuracy for muscle invasion in patients with BC before TURBT. Combining T2WI with dDLR might further improve the diagnostic accuracy of VI-RADS.