Detecting and grading prostate cancer in radical prostatectomy specimens through deep learning techniques

OBJECTIVES: This study aims to evaluate the ability of deep learning algorithms to detect and grade prostate cancer (PCa) in radical prostatectomy specimens. METHODS: We selected 12 whole-slide images of radical prostatectomy specimens. These images were divided into patches, and then, analyzed and annotated. The annotated areas were categorized as follows: stroma, normal glands, and Gleason patterns 3, 4, and 5. Two analyses were performed: i) a categorical image classification method that labels each image as benign or as Gleason 3, Gleason 4, or Gleason 5, and ii) a scanning method in which distinct areas representative of benign and different Gleason patterns are delineated and labeled separately by a pathologist. The Inception v3 Convolutional Neural Network architecture was used in categorical model training, and a Mask Region-based Convolutional Neural Network was used to train the scanning method. After training, we selected three new whole-slide images that were not used during the training to evaluate the model as our test dataset. The analysis results of the images using deep learning algorithms were compared with those obtained by the pathologists. RESULTS: In the categorical classification method, the trained model obtained a validation accuracy of 94.1% during training; however, the concordance with our expert uropathologists in the test dataset was only 44%. With the image-scanning method, our model demonstrated a validation accuracy of 91.2%. When the test images were used, the concordance between the deep learning method and uropathologists was 89%. CONCLUSION: Deep learning algorithms have a high potential for use in the diagnosis and grading of PCa. Scanning methods are likely to be superior to simple classification methods.

Comparative study of percutaneous nephrolithotomy performed in the traditional prone position and in three different supine positions

ABSTRACT Purpose: To compare the outcomes of percutaneous nephrolithotomy (PCNL) performed in the prone position (PRON) and in three variations of the supine position. Materials and Methods: We performed a retrospective analysis of patients that underwent PCNL at our institution from June 2011 to October 2016 in PRON and in three variations of the supine position: complete supine (COMPSUP), original Valdivia (VALD), and Galdakao - modified Valdivia (GALD). All patients had a complete pre - operative evaluation, including computed tomography (CT). Success was defined as the absence of residual fragments larger than 4 mm on the first post - operative day CT. Results: We analyzed 393 PCNLs: 100 in COMPSUP, 94 in VALD, 100 in GALD, and 99 in PRON. The overall success rate was 50.9% and was similar among groups (p = 0.428). There were no differences between groups in the number of punctures, stone - free rate, frequency of blood transfusions, drop in hemoglobin level, length of hospital stay, and severe complications (Clavien ≥ 3). COMPSUP had a significantly lower operative time than the other positions. COMPSUP had lower fluoroscopy time than VALD. Conclusion: Patient positioning in PCNL does not seem to impact the rates of success or severe complications. However, COMPSUP is associated with a shorter surgical time than the other positions.