RESUMEN
Recent advances in foundation models have revolutionized model development in digital pathology, reducing dependence on extensive manual annotations required by traditional methods. The ability of foundation models to generalize well with few-shot learning addresses critical barriers in adapting models to diverse medical imaging tasks. This work presents the Granular Box Prompt Segment Anything Model (GB-SAM), an improved version of the Segment Anything Model (SAM) fine-tuned using granular box prompts with limited training data. The GB-SAM aims to reduce the dependency on expert pathologist annotators by enhancing the efficiency of the automated annotation process. Granular box prompts are small box regions derived from ground truth masks, conceived to replace the conventional approach of using a single large box covering the entire H&E-stained image patch. This method allows a localized and detailed analysis of gland morphology, enhancing the segmentation accuracy of individual glands and reducing the ambiguity that larger boxes might introduce in morphologically complex regions. We compared the performance of our GB-SAM model against U-Net trained on different sizes of the CRAG dataset. We evaluated the models across histopathological datasets, including CRAG, GlaS, and Camelyon16. GB-SAM consistently outperformed U-Net, with reduced training data, showing less segmentation performance degradation. Specifically, on the CRAG dataset, GB-SAM achieved a Dice coefficient of 0.885 compared to U-Net's 0.857 when trained on 25% of the data. Additionally, GB-SAM demonstrated segmentation stability on the CRAG testing dataset and superior generalization across unseen datasets, including challenging lymph node segmentation in Camelyon16, which achieved a Dice coefficient of 0.740 versus U-Net's 0.491. Furthermore, compared to SAM-Path and Med-SAM, GB-SAM showed competitive performance. GB-SAM achieved a Dice score of 0.900 on the CRAG dataset, while SAM-Path achieved 0.884. On the GlaS dataset, Med-SAM reported a Dice score of 0.956, whereas GB-SAM achieved 0.885 with significantly less training data. These results highlight GB-SAM's advanced segmentation capabilities and reduced dependency on large datasets, indicating its potential for practical deployment in digital pathology, particularly in settings with limited annotated datasets.
RESUMEN
Background and Objective: Esophageal carcinoma with switch/sucrose nonfermenting (SWI/SNF)-related, matrix-associated, actin-dependent regulator of chromatin, subfamily A, member 4 (SMARCA4) mutation is a rare variant of malignant esophageal epithelial neoplasm, which is characterized by the loss of SMARCA4/BRG1 protein on immunohistochemistry or alterations in the SMARCA4 gene on sequencing. Only a few case series and case reports of esophageal carcinoma with SMARCA4 mutations have been published in the English literature; the rarity of the disease poses significant diagnostic challenges for surgical pathologists and could potentially lead to delayed or suboptimal patient care. Herein, we reviewed the available literature on esophageal carcinoma with SMARCA4 mutations to discuss its epidemiology, clinical presentation, pathological and molecular features, diagnostic challenges, treatment, and prognosis. Methods: The PubMed, Scopus, Ovid, and Google Scholar databases were extensively reviewed. The references included in the articles were cross-examined to identify any missing articles. We searched for all published literature on esophageal carcinoma with SMARCA4 mutations from inception of the databases to date. Key Content and Findings: Esophageal carcinoma with SMARCA4 mutations is most common in middle-aged and older men. Barrett esophagus and gastroesophageal reflux disease (GERD) are the most associated risk factors. Dysphagia was the most common initial clinical presentation. Esophagogastroduodenoscopy (EGD) is the preferred diagnostic modality. Microscopically, the tumor cells exhibited epithelioid features mixed with variable components of rhabdoid and glandular differentiation. The tumor cells showed variable immunoreactivity for cytokeratin and sometimes weakly expressed neuroendocrine or B-lymphocyte markers (Pax5), which are potential diagnostic pitfalls. Melanoma marker tests showed negative results. The SMARCB1/INI1 protein remains intact, and a definitive diagnosis necessitates the presence of either SMARCA4/BRG1 protein loss or SMARCA4 gene mutations. Esophageal carcinoma with SMARCA4 mutations shows overly aggressive behavior and presents with advanced stages of disease; most patients succumb to the disease within 1 year of initial diagnosis. Conclusions: Esophageal carcinoma with SMARCA4 mutation is an overly aggressive disease, and further research on the affected molecular pathway may help improve its prognosis.
RESUMEN
Existing natural language processing (NLP) methods to convert free-text clinical notes into structured data often require problem-specific annotations and model training. This study aims to evaluate ChatGPT's capacity to extract information from free-text medical notes efficiently and comprehensively. We developed a large language model (LLM)-based workflow, utilizing systems engineering methodology and spiral "prompt engineering" process, leveraging OpenAI's API for batch querying ChatGPT. We evaluated the effectiveness of this method using a dataset of more than 1000 lung cancer pathology reports and a dataset of 191 pediatric osteosarcoma pathology reports, comparing the ChatGPT-3.5 (gpt-3.5-turbo-16k) outputs with expert-curated structured data. ChatGPT-3.5 demonstrated the ability to extract pathological classifications with an overall accuracy of 89%, in lung cancer dataset, outperforming the performance of two traditional NLP methods. The performance is influenced by the design of the instructive prompt. Our case analysis shows that most misclassifications were due to the lack of highly specialized pathology terminology, and erroneous interpretation of TNM staging rules. Reproducibility shows the relatively stable performance of ChatGPT-3.5 over time. In pediatric osteosarcoma dataset, ChatGPT-3.5 accurately classified both grades and margin status with accuracy of 98.6% and 100% respectively. Our study shows the feasibility of using ChatGPT to process large volumes of clinical notes for structured information extraction without requiring extensive task-specific human annotation and model training. The results underscore the potential role of LLMs in transforming unstructured healthcare data into structured formats, thereby supporting research and aiding clinical decision-making.
RESUMEN
INTRODUCTION: Differentiating pancreatic serous cystadenoma (SCA) from well-differentiated neuroendocrine tumors (WDNETs) based on histomorphology is critical yet challenging, particularly in small biopsy samples. Our study aimed to examine the expression profile of INSM1 in cytologic and surgical resection specimens from pancreatic SCA to evaluate its potential as a discriminative marker against pancreatic WDNET. METHODS: We characterized INSM1 immunohistochemistry in 34 patients with pancreatic SCA, comprising 23 surgical resections and 11 cytology specimens. As a control, we used 28 cytology specimens from pancreatic WDNET. Clinical information was retrieved through a review of electronic medical records. RESULTS: All 11 pancreatic SCA cytology specimens and 15 of 23 pancreatic SCA surgical resections exhibited absent INSM1 immunostaining. Each of the remaining eight surgical resection specimens demonstrated 1 % immunoreactivity. In contrast, 27 out of 28 (96 %) pancreatic WDNET cytology specimens were positive for INSM1 immunostaining, with a median immunoreactivity of 90 % and a range of 30-90 %. Overall, INSM1 immunostains perform similarly to chromogranin and synaptophysin in pancreatic SCA. CONCLUSIONS: The results indicate that INSM1 immunohistochemistry staining may serve as a useful neuroendocrine marker to differentiate pancreatic SCA from pancreatic WDNET in clinical practice. To our knowledge, this represents the first large-scale study to evaluate INSM1 immunostaining in surgical and cytology specimens from pancreatic SCA.