Histopathological assessment of transplant biopsies is currently the standard method to diagnose allograft rejection and can help guide patient management, but it is one of the most challenging areas of pathology, requiring considerable expertise, time, and effort. We aimed to analyse the utility of deep learning to preclassify histology of kidney allograft biopsies into three main broad categories (ie, normal, rejection, and other diseases) as a potential biopsy triage system focusing on transplant rejection. We performed a retrospective, multicentre, proof-of-concept study using 5844 digital whole slide images of kidney allograft biopsies from 1948 patients. Kidney allograft biopsy samples were identified by a database search in the Departments of Pathology of the Amsterdam UMC, Amsterdam, Netherlands (1130 patients) and the University Medical Center Utrecht, Utrecht, Netherlands (717 patients). 101 consecutive kidney transplant biopsies were identified in the archive of the Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany. Convolutional neural networks (CNNs) were trained to classify allograft biopsies as normal, rejection, or other diseases. Three times cross-validation (1847 patients) and deployment on an external real-world cohort (101 patients) were used for validation. Area under the receiver operating characteristic curve (AUROC) was used as the main performance metric (the primary endpoint to assess CNN performance). Serial CNNs, first classifying kidney allograft biopsies as normal (AUROC 0·87 [ten times bootstrapped CI 0·85-0·88]) and disease (0·87 [0·86-0·88]), followed by a second CNN classifying biopsies classified as disease into rejection (0·75 [0·73-0·76]) and other diseases (0·75 [0·72-0·77]), showed similar AUROC in cross-validation and deployment on independent real-world data (first CNN normal AUROC 0·83 [0·80-0·85], disease 0·83 [0·73-0·91]; second CNN rejection 0·61 [0·51-0·70], other diseases 0·61 [0·50-0·74]). A single CNN classifying biopsies as normal, rejection, or other diseases showed similar performance in cross-validation (normal AUROC 0·80 [0·73-0·84], rejection 0·76 [0·66-0·80], other diseases 0·50 [0·36-0·57]) and generalised well for normal and rejection classes in the real-world data. Visualisation techniques highlighted rejection-relevant areas of biopsies in the tubulointerstitium. This study showed that deep learning-based classification of transplant biopsies could support pathological diagnostics of kidney allograft rejection. European Research Council; German Research Foundation; German Federal Ministries of Education and Research, Health, and Economic Affairs and Energy; Dutch Kidney Foundation; Human(e) AI Research Priority Area of the University of Amsterdam; and Max-Eder Programme of German Cancer Aid.

Head and neck squamous cell carcinoma (HNSCC) is an immunogenic cancer type, and tumor associated macrophages (TAMs) are a major component of the tumor microenvironment (TME). In this systematic review and meta-analysis, studies assessing tumor infiltration with CD68+, iNOS+, HLA-DR+, CD11b+, CD163+, CD206+, and CD204+TAMs were included, and correlation to survival hazard was studied. A low number of CD68+TAMs correlated to better overall survival (OS) in multivariate analysis (HR 1.36 95 %CI (1.07-1.72) P = .01). CD68+TAMs did not correlate to disease free survival (DFS), disease specific survival (DSS), progression free survival (PFS), or recurrence free survival (RFS). A low number of CD163+TAMs correlated to better OS in uni- and multivariate analysis (resp. HR 2.65 95 %CI (1.57-4.46) P = .01 and HR 2.42 95 %CI (1.72-3.41) P < .001). A low number of CD163+TAMs also correlated to better DFS and PFS, whereas a low number of CD204+TAMs only correlated to PFS. While IHC analysis of pan macrophage marker CD68 and M2-like marker CD163 both show prognostic utility in OS, CD163 is a stronger prognosticator, as indicated by multivariate meta-analysis. CD163+TAMs also correlate to DFS and PFS; outcomes that are more relevant to patients, thus showing promising results for future clinical implementation.

OBJECT: Endoscopic third ventriculostomy (ETV) is an effective but technically demanding procedure with significant risk. Current simulators, including human cadavers, animal models, and virtual reality systems, are expensive, relatively inaccessible, and can lack realistic sensory feedback. The purpose of this study was to construct a realistic, low-cost, reusable brain simulator for ETV and evaluate its fidelity. METHODS: A brain silicone replica mimicking normal mechanical properties of a 4-month-old child with hydrocephalus was constructed, encased in the replicated skull, and immersed in water. Realistic intraventricular landmarks included the choroid plexus, veins, mammillary bodies, infundibular recess, and basilar artery. The thinned-out third ventricle floor, which dissects appropriately, is quickly replaceable. Standard neuroendoscopic equipment including irrigation is used. Bleeding scenarios are also incorporated. A total of 16 neurosurgical trainees (Postgraduate Years 1–6) and 9 pediatric and adult neurosurgeons tested the simulator. All participants filled out questionnaires (5-point Likert-type items) to rate the simulator for face and content validity. RESULTS: The simulator is portable, robust, and sets up in minutes. More than 95% of participants agreed or strongly agreed that the simulator's anatomical features, tissue properties, and bleeding scenarios were a realistic representation of that seen during an ETV. Participants stated that the simulator helped develop the required hand-eye coordination and camera skills, and the training exercise was valuable. CONCLUSIONS: A low-cost, reusable, silicone-based ETV simulator realistically represents the surgical procedure to trainees and neurosurgeons. It can help them develop the technical and cognitive skills for ETV including dealing with complications.

The rapid introduction of digital pathology has greatly facilitated development of artificial intelligence (AI) models in pathology that have shown great promise in assisting morphological diagnostics and quantitation of therapeutic targets. We are now at a tipping point where companies have started to bring algorithms to the market, and questions arise whether the pathology community is ready to implement AI in routine workflow. However, concerns also arise about the use of AI in pathology. This article reviews the pros and cons of introducing AI in diagnostic pathology.

Background: Melanoma is an aggressive form of skin cancer in which tumor-infiltrating lymphocytes (TILs) are a biomarker for recurrence and treatment response. Manual TIL assessment is prone to interobserver variability, and current deep learning models are not publicly accessible or have low performance. Deep learning models, however, have the potential of consistent spatial evaluation of TILs and other immune cell subsets with the potential of improved prognostic and predictive value. To make the development of these models possible, we created the Panoptic Segmentation of nUclei and tissue in advanced MelanomA (PUMA) dataset and assessed the performance of several state-of-the-art deep learning models. In addition, we show how to improve model performance further by using heuristic postprocessing in which nuclei classes are updated based on their tissue localization. Results: The PUMA dataset includes 155 primary and 155 metastatic melanoma hematoxylin and eosin–stained regions of interest with nuclei and tissue annotations from a single melanoma referral institution. The Hover-NeXt model, trained on the PUMA dataset, demonstrated the best performance for lymphocyte detection, approaching human interobserver agreement. In addition, heuristic postprocessing of deep learning models improved the detection of noncommon classes, such as epithelial nuclei. Conclusion: The PUMA dataset is the first melanoma-specific dataset that can be used to develop melanoma-specific nuclei and tissue segmentation models. These models can, in turn, be used for prognostic and predictive biomarker development. Incorporating tissue and nuclei segmentation is a step toward improved deep learning nuclei segmentation performance. To support the development of these models, this dataset is used in the PUMA challenge.