AI Model for Prostate Biopsies Predicts Cancer Survival

Abstract

An artificial intelligence (AI) algorithm for prostate cancer detection and grading was developed for clinical diagnostics on biopsies. The study cohort included 4221 scanned slides from 872 biopsy sessions at the HUS Helsinki University Hospital during 2016–2017 and a subcohort of 126 patients treated by robot-assisted radical prostatectomy (RALP) during 2016–2019. In the validation cohort (n = 391), the model detected cancer with a sensitivity of 98% and specificity of 98% (weighted kappa 0.96 compared with the pathologist’s diagnosis). Algorithm-based detection of the grade area recapitulated the pathologist’s grade group. The area of AI-detected cancer was associated with extra-prostatic extension (G5 OR: 48.52; 95% CI 1.11–8.33), seminal vesicle invasion (cribriform G4 OR: 2.46; 95% CI 0.15–1.7; G5 OR: 5.58; 95% CI 0.45–3.42), and lymph node involvement (cribriform G4 OR: 2.66; 95% CI 0.2–1.8; G5 OR: 4.09; 95% CI 0.22–3). Algorithm-detected grade group 3–5 prostate cancer depicted increased risk for biochemical recurrence compared with grade groups 1–2 (HR: 5.91; 95% CI 1.96–17.83). This study showed that a deep learning model not only can find and grade prostate cancer on biopsies comparably with pathologists but also can predict adverse staging and probability for recurrence after surgical treatment.

Deep Learning Methodologies Applied to Digital Pathology in Prostate Cancer: A Systematic Review

“…These articles achieve an AUC of 0.99 on internal validation cohorts and above 0.90 on external cohorts. Indeed, many other approaches obtained AUC over 0.95 [ 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ]. In 2021, the FDA approved the deployment of PaigeProstate [ 34 ], based on the MIL architecture of Campanella [ 32 ].…”

“…DeepDx was further evaluated on an external cohort by Jung et al [ 75 ] (kappa of 0.65 and quadratic kappa of 0.90). A third algorithm capable of Gleason grading (kappa of 0.77) was commercialized by Aiforia [ 42 ]. Another important milestone in the field was the organization and the release of the PANDA challenge focusing on Gleason grading at the WSI level without gland annotations [ 76 ].…”

“…When focusing on recurrence, AUC around 0.8 [ 80 , 81 ] and Hazard Ratios (HR) above 4.8 [ 42 , 82 , 83 ] were obtained. A couple of articles studied the probability of developing metastasis [ 84 , 85 ].…”

Applications of artificial intelligence in urologic oncology

“…Each study’s inclusion and exclusion criteria were highly heterogeneous, but they all focused on urological cancer. Finally, 58 studies on urological cancers (prostate cancer: 21 [ 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 ], bladder cancer: 20 [ 5 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 ], kidney cancer: 17 [ 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 ]) were identified as shown in Fig. 1 .…”

Harnessing artificial intelligence for prostate cancer management

“… 17 Aiforia Clinical AI Model for Prostate Cancer Aiforia cancer detection and Gleason grading Finland Sandeman et al. 18 DeepDx-Prostate Connect Deep Bio ROI detection, Gleason grading, and quantification South Korea Jung et al. 19 , 20 and Ryu et al.…”