Figure 2.

Image acquisition workflow using the automated day optimizing throughput (DOT) engine and biparametric imaging (in orange) as well as the deep learning architecture with the preprocessing pipeline (in blue) and the deep-learning-based lesion detection and classification component (in green). DOT = day optimizing throughput, FOV = field of view, 3D = three-dimensional, TA = time of acquisition, DICOM = Digital Imaging and Communications in Medicine, ADC = apparent diffusion coefficient, FP = false positive, PI-RADS = Prostate Imaging Reporting and Data System.