Prediction of invasive cribriform carcinoma based on biparametric MRI in the patients with clinically significant prostate cancer

Motivation: Invasive cribriform carcinoma (CRC) of the prostate is frequently missed on biopsy. Additionally, the visibility of CRC on MRI has shown inconsistent results in previous studies, and MRI findings specific to CRC remain under-explored. Goal(s): To identify clinical and MRI features that may help predict the presence of CRC. Approach: This study analyzed clinical parameters and MRI findings in 348 patients with surgically confirmed clinically significant prostate cancer (csPCa). Results: High PSA density, low ADC ratio, multizonal involvement, and extraprostatic extension were significant indicators for identifying CRC. A logistic regression model achieved an AUC of 0.71 in predicting CRC among csPCa cases. Impact: Accurately predicting CRC through pre-operative imaging could potentially improve management strategies for csPCa patients. The imaging features identified are straightforward to apply in clinical settings, and advanced methods, such as radiomics, may further enhance diagnostic accuracy.