RESUMEN
In our rapidly expanding landscape of artificial intelligence, synthetic data have become a topic of great promise and also some concern. This review aimed to provide pathologists and laboratory professionals with a primer on the role of synthetic data and how it may soon shape the landscape within our field. Using synthetic data presents many advantages but also introduces a milieu of new obstacles and limitations. This review aimed to provide pathologists and laboratory professionals with a primer on the general concept of synthetic data and its potential to transform our field. By leveraging synthetic data, we can help accelerate the development of various machine learning models and enhance our medical education and research/quality study needs. This review explored the methods for generating synthetic data, including rule-based, machine learning model-based and hybrid approaches, as they apply to applications within pathology and laboratory medicine. We also discussed the limitations and challenges associated with such synthetic data, including data quality, malicious use, and ethical bias/concerns and challenges. By understanding the potential benefits (ie, medical education, training artificial intelligence programs, and proficiency testing, etc) and limitations of this new data realm, we can not only harness its power to improve patient outcomes, advance research, and enhance the practice of pathology but also become readily aware of their intrinsic limitations.
Asunto(s)
Aprendizaje Automático , Humanos , Patología , Inteligencia ArtificialRESUMEN
Certain features are helpful in the identification of gunshot entrance and exit wounds, such as the presence of muzzle imprints, peripheral tears, stippling, bone beveling, and wound border irregularity. Some cases are less straightforward and wounds can thus pose challenges to an emergency room doctor or forensic pathologist. In recent years, deep learning has shown promise in various automated medical image classification tasks. This study explores the feasibility of using a deep learning model to classify entry and exit gunshot wounds in digital color images. A collection of 2418 images of entrance and exit gunshot wounds were procured. Of these, 2028 entrance and 1314 exit wounds were cropped, focusing on the area around each gunshot wound. A ConvNext Tiny deep learning model was trained using the Fastai deep learning library, with a train/validation split ratio of 70/30, until a maximum validation accuracy of 92.6% was achieved. An additional 415 entrance and 293 exit wound images were collected for the test (holdout) set. The model achieved an accuracy of 87.99%, precision of 83.99%, recall of 87.71%, and F1-score 85.81% on the holdout set. Correctly classified were 88.19% of entrance wounds and 87.71% of exit wounds. The results are comparable to what a forensic pathologist can achieve without other morphologic cues. This study represents one of the first applications of artificial intelligence to the field of forensic pathology. This work demonstrates that deep learning models can discern entrance and exit gunshot wounds in digital images with high accuracy.
RESUMEN
BACKGROUND: The Ki-67 index is important for grading neuroendocrine tumors (NETs) in cytology. However, different counting methods exist. Recently, augmented reality microscopy (ARM) has enabled real-time image analysis using glass slides. The objective of the current study was to compare different traditional Ki-67 scoring methods in cell block material with newer methods such as ARM. METHODS: Ki-67 immunostained slides from 50 NETs of varying grades were retrieved (39 from the pancreas and 11 metastases). Methods with which to quantify the Ki-67 index in up to 3 hot spots included: 1) "eyeball" estimation (EE); 2) printed image manual counting (PIMC); 3) ARM with live image analysis; and 4) image analysis using whole-slide images (WSI) (field of view [FOV] and the entire slide). RESULTS: The Ki-67 index obtained using the different methods varied. The pairwise kappa results varied from no agreement for image analysis using digital image analysis WSI (FOV) and histology to near-perfect agreement for ARM and PIMC. Using surgical pathology as the gold standard, the EE method was found to have the highest concordance rate (84.2%), followed by WSI analysis of the entire slide (73.7%) and then both the ARM and PIMC methods (63.2% for both). The PIMC method was the most time-consuming whereas image analysis using WSI (FOV) was the fastest method followed by ARM. CONCLUSIONS: The Ki-67 index for NETs in cell block material varied by the method used for scoring, which may affect grade. PIMC was the most time-consuming method, and EE had the highest concordance rate. Although real-time automated counting using image analysis demonstrated inaccuracies, ARM streamlined and hastened the task of Ki-67 quantification in NETs.
Asunto(s)
Realidad Aumentada , Biomarcadores de Tumor/análisis , Proliferación Celular , Procesamiento de Imagen Asistido por Computador/métodos , Antígeno Ki-67/análisis , Microscopía/métodos , Tumores Neuroendocrinos/patología , Humanos , Neoplasias Hepáticas/secundario , Clasificación del Tumor , Tumores Neuroendocrinos/inmunología , Neoplasias Pancreáticas/inmunología , Neoplasias Pancreáticas/patologíaRESUMEN
Sudden cardiac death is an unexpected clinical condition that typically occurs due to a cardiac cause, generally within 1â¯h of symptom onset, in people with known or unknown cardiac disease. Primary malignant pericardial mesothelioma, as a cause of sudden death, is an uncommon consequence of a rare disease. Herein, we present a case of cardiac tamponade due to a primary pericardial mesothelioma. Cytological, histopathology and gross post-mortem findings, in a previously asymptomatic 46-old-year man, are reported. The medical literature regarding this topic is also reviewed.
RESUMEN
INTRODUCTION: Arborizing stromal meshwork fragments (ASMFs) have been proposed as a useful clue to differentiate mucin associated with mucinous adenocarcinoma from contaminating mucus during endobronchial ultrasound-guided transbronchial needle aspiration. Our aim was to retrospectively review cytology cases with mucinous material to determine the utility of ASMFs in diagnosing mucinous tumors. MATERIAL AND METHODS: Diff-Quik stained smears from archival cytology cases (N = 40) were reviewed, including adenocarcinomas with mucinous features, cystic mucinous neoplasms, and control cases with mucin contamination. Specimens were procured by image-guided fine needle aspiration (FNA) (16 cases), endoscopic ultrasound-guided FNA (22 cases), pathologist-performed FNA (1 case), and fluid drainage (1 case). All cases were reviewed for ASMFs, which were defined as metachromatic, spidery extensions with frayed edges within a background of mucinous material. RESULTS: ASMFs were identified in 4 (10% of cases, 14% of adenocarcinomas) cases of metastatic gastrointestinal mucinous adenocarcinomas in various locations (liver, lymph node, lung, and bone), but absent in mucin contamination. ASMFs in Diff-Quik stained smears were magenta-colored and corresponded to intervening stroma between dissecting mucin in the tumor. Nonarborizing desmoplastic stroma, inspissated mucus, cartilage fragments, transgressing vessels in renal cell carcinoma, and mucus-like material in pancreatic pseudocysts can morphologically mimic ASMFs. CONCLUSION: These data show that ASMFs may be encountered in some (14%) cases of adenocarcinoma with mucinous differentiation. When present, ASMFs can be diagnostically helpful to differentiate adenocarcinoma with mucinous features from contaminating mucus, if reliably distinguished from mimics. Diagn. Cytopathol. 2016;44:805-810. © 2016 Wiley Periodicals, Inc.