Hybrid T2-weighted and diffusion-weighted magnetic resonance imaging for differentiating prostate cancer from benign prostatic hyperplasia

The objective of our study was to evaluate the role of a hybrid T2-weighted and diffusion-weighted imaging for the differentiation of benign prostatic hyperplasia (BPH) and prostate cancer (PCa). Eight patients with PCa who underwent preoperative 3-T magnetic resonance imaging (MRI) and prostatectomy were included in this study. Patients underwent a hybrid T2-weighted and diffusion-weighted imaging examination consisting of diffusion-weighted images acquired with the echo times (TEs) of 47, 75 and 100 ms and b-values of 0, 750 and 1500 s/mm2. The apparent diffusion coefficient (ADC), T2 constant and signal intensity of DWI at the highest b-value were selected as features separately or combined with either linear discriminant analysis (LDA) or quadratic discriminant analysis (QDA) classifier to differentiate BPH and PCa. Receiver operating characteristic analysis and leave-one-out cross validation were used to evaluate the accuracy of the classification. The maximum area under curve (AUC) of the ADC, T2 and signal intensity of DWI in classifying BPH and PCa were 0.80 ± 0.07, 0.78 ± 0.09 and 0.82 ± 0.07, respectively. The maximum AUC of the LDA or QDA classifier output in the same task was 0.93 ± 0.04 or 0.98 ± 0.02, respectively. Thus, hybrid MRI measures the response of ADC and T2 to changing TEs and b-values, respectively. This approach combines ADC at different TEs and T2 at different b-values and is effective in designing the classifier to differentiate BPH and PCa tissue.