Search

Prostate cancer is a significant cause of morbidity and mortality in the USA. In this paper, we develop a computer-aided diagnostic (CAD) system for automated grade groups (GG) classification using digitized prostate biopsy specimens (PBSs). Our CAD system aims to firstly classify the Gleason pattern (GP), and then identifies the Gleason score (GS) and GG. The GP classification pipeline is based on a pyramidal deep learning system that utilizes three convolution neural networks (CNN) to produce both patch- and pixel-wise classifications. The analysis starts with sequential preprocessing steps that include a histogram equalization step to adjust intensity values, followed by a PBSs’ edge enhancement. The digitized PBSs are then divided into overlapping patches with the three sizes: 100 × 100 (CNNS), 150 × 150 (CNNM), and 200 × 200 (CNNL), pixels, and 75% overlap. Those three sizes of patches represent the three pyramidal levels. This pyramidal technique allows us to extract rich information, such as that the larger patches give more global information, while the small patches provide local details. After that, the patch-wise technique assigns each overlapped patch a label as GP categories (1 to 5). Then, the majority voting is the core approach for getting the pixel-wise classification that is used to get a single label for each overlapped pixel. The results after applying those techniques are three images of the same size as the original, and each pixel has a single label. We utilized the majority voting technique again on those three images to obtain only one. The proposed framework is trained, validated, and tested on 608 whole slide images (WSIs) of the digitized PBSs. The overall diagnostic accuracy is evaluated using several metrics: precision, recall, F1-score, accuracy, macro-averaged, and weighted-averaged. The (CNNL) has the best accuracy results for patch classification among the three CNNs, and its classification accuracy is 0.76. The macro-averaged and weighted-average metrics are found to be around 0.70–0.77. For GG, our CAD results are about 80% for precision, and between 60% to 80% for recall and F1-score, respectively. Also, it is around 94% for accuracy and NPV. To highlight our CAD systems’ results, we used the standard ResNet50 and VGG-16 to compare our CNN’s patch-wise classification results. As well, we compared the GG’s results with that of the previous work.

Abstract Background A new 5-tiered grading grouping system has recently been endorsed for reporting of prostate cancer (PCa) grade to better reflect escalating risk of progression and cancer death. While several validations of the new grade groupings have been undertaken, most have involved centralised pathological review by specialist urological pathologists. Methods Participants included 4268 men with non-metastatic PCa diagnosed between 2006 and 2013 from the multi-institutional South Australia Prostate Cancer Clinical Outcomes Collaborative registry. PCa-specific survival and biochemical recurrence-free survival were compared across the five grade groups using multivariable competing risk regression. Results For the entire cohort, risk of PCa death increased with increasing grade groups (at biopsy) Adjusted subdistribution-hazard ratios [sHR] and 95% confidence intervals [95%CI] were: 2.2 (1.5–3.6); 2.5 (1.6–4.2); 4.1 (2.6–6.7) and 8.7 (4.5–14.0) for grade groups II (pattern 3 + 4), III (pattern 4 + 3), IV (total score 8) and V (total score 9–10) respectively, relative to grade group I (total score

Background: Observational data and funnel plots are routinely used outside of pathology to understand trends and improve performance. Objective: Extract diagnostic rate (DR) information from free text surgical pathology reports with synoptic elements and assess whether inter-rater variation and clinical history completeness information useful for continuous quality improvement (CQI) can be obtained. Methods: All in-house prostate biopsies in a 6-year period at two large teaching hospitals were extracted and then diagnostically categorized using string matching, fuzzy string matching, and hierarchical pruning. DRs were then stratified by the submitting physicians and pathologists. Funnel plots were created to assess for diagnostic bias. Results: 3,854 prostate biopsies were found and all could be diagnostically classified. Two audits involving the review of 700 reports and a comparison of the synoptic elements with the free text interpretations suggest a categorization error rate of 40 cases and together assessed 3,690 biopsies. There was considerable inter-rater variability and a trend toward more World Health Organization/International Society of Urologic Pathology Grade 1 cancers in older pathologists. Normalized deviations plots, constructed using the median DR, and standard error can elucidate associated over- and under-calls for an individual pathologist in relation to their practice group. Clinical history completeness by submitting medical doctor varied significantly (100% to 22%). Conclusion: Free text data analyses have some limitations; however, they could be used for data-driven CQI in anatomical pathology, and could lead to the next generation in quality of care.