Development of a Deep Learning Algorithm for the Histopathologic Diagnosis and Gleason Grading of Prostate Cancer Biopsies: A Pilot Study

• Published in: European urology focus Vol. 7; no. 2; pp. 347 - 351
• Main Authors: Kott, Ohad, Linsley, Drew, Amin, Ali, Karagounis, Andreas, Jeffers, Carleen, Golijanin, Dragan, Serre, Thomas, Gershman, Boris
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.03.2021; Elsevier
• Subjects: Algorithms; Artificial Intelligence; Biopsy; Computer Science; Deep Learning; Diagnosis; Gleason grade; Human health and pathology; Humans; Image Interpretation, Computer-Assisted; Life Sciences; Machine learning; Male; Neoplasm Grading; Pilot Projects; Prostate - pathology; Prostate cancer; Prostatic Neoplasms - pathology; Deep learning; Gleason grade; Diagnosis; Prostate cancer; Machine learning
• ISSN: 2405-4569; 2405-4569
• DOI: 10.1016/j.euf.2019.11.003

The pathologic diagnosis and Gleason grading of prostate cancer are time-consuming, error-prone, and subject to interobserver variability. Machine learning offers opportunities to improve the diagnosis, risk stratification, and prognostication of prostate cancer. To develop a state-of-the-art deep learning algorithm for the histopathologic diagnosis and Gleason grading of prostate biopsy specimens. A total of 85 prostate core biopsy specimens from 25 patients were digitized at 20× magnification and annotated for Gleason 3, 4, and 5 prostate adenocarcinoma by a urologic pathologist. From these virtual slides, we sampled 14803 image patches of 256×256 pixels, approximately balanced for malignancy. We trained and tested a deep residual convolutional neural network to classify each patch at two levels: (1) coarse (benign vs malignant) and (2) fine (benign vs Gleason 3 vs 4 vs 5). Model performance was evaluated using fivefold cross-validation. Randomization tests were used for hypothesis testing of model performance versus chance. The model demonstrated 91.5% accuracy (p<0.001) at coarse-level classification of image patches as benign versus malignant (0.93 sensitivity, 0.90 specificity, and 0.95 average precision). The model demonstrated 85.4% accuracy (p<0.001) at fine-level classification of image patches as benign versus Gleason 3 versus Gleason 4 versus Gleason 5 (0.83 sensitivity, 0.94 specificity, and 0.83 average precision), with the greatest number of confusions in distinguishing between Gleason 3 and 4, and between Gleason 4 and 5. Limitations include the small sample size and the need for external validation. In this study, a deep learning-based computer vision algorithm demonstrated excellent performance for the histopathologic diagnosis and Gleason grading of prostate cancer. We developed a deep learning algorithm that demonstrated excellent performance for the diagnosis and grading of prostate cancer. In this pilot study, a deep learning-based computer vision algorithm demonstrated excellent accuracy for the histopathologic diagnosis and Gleason grading of prostate cancer. These results are encouraging for the future clinical application of automated histopathologic diagnosis via deep learning.