medRxiv - Pathology最新文献

{"title":"Clinical-Grade Validation of an Autofluorescence Virtual Staining System with Human Experts and a Deep Learning System for Prostate Cancer","authors":"Pok Fai Wong, Carson McNeil, Yang Wang, Jack Paparian, Charles Santori, Michael Gutierrez, Andrew Homyk, Kunal Nagpal, Tiam Jaroensri, Ellery Wulczyn, Julia Sigman, David Steiner, Sudha Rao, Po-Hsuan Cameron Cheng, Luke Restoric, Jonathan Roy, Peter Cimermancic","doi":"10.1101/2024.03.27.24304447","DOIUrl":"https://doi.org/10.1101/2024.03.27.24304447","url":null,"abstract":"The tissue diagnosis of adenocarcinoma and intraductal carcinoma of the prostate (IDC-P) includes Gleason grading of tumor morphology on the hematoxylin and eosin (H&E) stain, and immunohistochemistry (IHC) markers on the PIN-4 stain (CK5/6, P63, AMACR). In this work, we create an automated system for producing both virtual H&E and PIN-4 IHC stains from unstained prostate tissue using a high-throughput multispectral fluorescence microscope and artificial intelligence & machine learning. We demonstrate that the virtual stainer models can produce high-quality images suitable for diagnosis by genitourinary pathologists. Specifically, we validate our system through extensive human review and computational analysis, using a previously-validated Gleason scoring model, and an expert panel, on a large dataset of test slides. This study extends our previous work on virtual staining from autofluorescence, demonstrates the clinical utility of this technology for prostate cancer, and exemplifies a rigorous standard of qualitative and quantitative evaluation for digital pathology.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"111 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140324049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"3D imaging of protein aggregates in human neurodegeneration by multiscale X-ray phase-contrast tomography","authors":"Jakob Reichmann, Jonas Franz, Marina Eckermann, Christine Stadelmann, Tim Salditt","doi":"10.1101/2024.03.26.24304193","DOIUrl":"https://doi.org/10.1101/2024.03.26.24304193","url":null,"abstract":"This study leverages X-ray phase-contrast tomography (XPCT) for detailed analysis of neurodegenerative diseases like Alzheimer's and Parkinson's, focusing on the 3D visualization and quantification of neuropathological features within fixed human postmortem tissue. XPCT, utilizing synchrotron radiation, offers micrometer resolution, enabling the examination of intraneuronal aggregates-Lewy bodies, granulovacuolar degeneration, Hirano bodies, and neurofibrillary tangles-and extracellular amyloid plaques and cerebral amyloid angiopathy in the fixed human tissue. This approach surpasses aspects of traditional histology, integrating with neuropathology workflows for the identification and high-resolution study of these features. It facilitates correlative studies and quantitative electron density assessments, providing insights into the structural composition and distribution of neurodegenerative pathologies.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"18 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140316711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generating highly accurate pathology reports from gigapixel whole slide images with HistoGPT","authors":"Manuel Tran, Paul Schmidle, Sophia J. Wagner, Valentin Koch, Valerio Lupperger, Annette Feuchtinger, Alexander Boehner, Robert Kaczmarczyk, Tilo Biedermann, Kilian Eyerich, Stephan A. Braun, Tingying Peng, Carsten Marr","doi":"10.1101/2024.03.15.24304211","DOIUrl":"https://doi.org/10.1101/2024.03.15.24304211","url":null,"abstract":"Histopathology is considered the gold standard for determining the presence and nature of disease, particularly cancer. However, the process of analyzing tissue samples and producing a final pathology report is time-consuming, labor-intensive, and non-standardized. Therefore, new technological solutions are being sought to reduce the workload of pathologists. In this work, we present HistoGPT, a vision language model that takes digitized slides as input and generates reports that match the quality of human-written reports, as confirmed by natural language processing metrics and domain expert evaluations. We show that HistoGPT generalizes to five international cohorts and can predict tumor subtypes and tumor thickness in a zero-shot fashion. Our work represents an important step toward integrating AI into the medical workflow. We publish both model code and weights so that the scientific community can apply and improve HistoGPT to advance the field of computational pathology.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"31 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140172242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital Pathology, Deep Learning, and Cancer: A Narrative Review","authors":"Darnell K Adrian Williams, Gillian Graifman, Nowair Hussain, Maytal Amiel, Tran Priscilla, Arjun Reddy, Ali Haider, Bali Kumar Kavitesh, Austin Li, Leael Alishahian, Nichelle Perera, Corey Efros, Myoungmee Babu, Mathew Tharakan, Mill Etienne, Benson Babu","doi":"10.1101/2024.03.14.24304308","DOIUrl":"https://doi.org/10.1101/2024.03.14.24304308","url":null,"abstract":"Background and Objective: Cancer is a leading cause of morbidity and mortality worldwide. The emergence of digital pathology and deep learning technologies signifies a transformative era in healthcare. These technologies can enhance cancer detection, streamline operations, and bolster patient care. A substantial gap exists between the development phase of deep learning models in controlled laboratory environments and their translations into clinical practice. This narrative review evaluates the current landscape of deep learning and digital pathology, analyzing the factors influencing model development and implementation into clinical practice.\u0000Methods: We searched multiple databases, including Web of Science, Arxiv, MedRxiv, BioRxiv, Embase, PubMed, DBLP, Google Scholar, IEEE Xplore, and Cochrane, targeting articles on whole slide imaging and deep learning published from 2014 and 2023. Out of 776 articles identified based on inclusion criteria, we selected 36 papers for the analysis.\u0000Key Content and Findings: Most articles in this review focus on the in-laboratory phase of deep learning model development, a critical stage in the deep learning lifecycle. Challenges arise during model development and their integration into clinical practice. Notably, lab performance metrics may not always match real-world clinical outcomes. As technology advances and regulations evolve, we expect more clinical trials to bridge this performance gap and validate deep learning models' effectiveness in clinical care. High clinical accuracy is vital for informed decision-making throughout a patient's cancer care.\u0000Conclusions: Deep learning technology can enhance cancer detection, clinical workflows, and patient care. Challenges may arise during model development. The deep learning lifecycle involves data preprocessing, model development, and clinical implementation. Achieving health equity requires including diverse patient groups and eliminating bias during implementation. While model development is integral, most articles focus on the pre-deployment phase. Future longitudinal studies are crucial for validating models in real-world settings post-deployment. A collaborative approach among computational pathologists, technologists, industry, and healthcare providers is essential for driving adoption in clinical settings.\u0000Keywords: Artificial Intelligence, Deep Learning, Digital Pathology, Computational Pathology, Cancer","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"51 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140153159","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of ChatGPT's Usefulness and Accuracy in Diagnostic Surgical Pathology.","authors":"Vincenzo Guastafierro, Devin Nicole Corbitt, Alessandra Bressan, Bethania Fernandes, Ömer Mintemur, Francesca Magnoli, Susanna Ronchi, Stefano La Rosa, Silvia Uccella, Salvatore Lorenzo Renne","doi":"10.1101/2024.03.12.24304153","DOIUrl":"https://doi.org/10.1101/2024.03.12.24304153","url":null,"abstract":"ChatGPT is an artificial intelligence capable of processing and generating human-like language. ChatGPT's role within clinical patient care and medical education has been explored; however, assessment of its potential in supporting histopathological diagnosis is lacking. In this study, we assessed ChatGPT's reliability in addressing pathology-related diagnostic questions across 10 subspecialties, as well as its ability to provide scientific references. We created five clinico-pathological scenarios for each subspecialty, posed to ChatGPT as open-ended or multiple-choice questions. Each question either asked for scientific references or not. Outputs were assessed by six pathologists according to: 1) usefulness in supporting the diagnosis and 2) absolute number of errors. All references were manually verified. We used directed acyclic graphs and structural causal models to determine the effect of each scenario type, field, question modality and pathologist evaluation. Overall, we yielded 894 evaluations. ChatGPT provided useful answers in 62.2% of cases. 32.1% of outputs contained no errors, while the remaining contained at least one error (maximum 18). ChatGPT provided 214 bibliographic references: 70.1% were correct, 12.1% were inaccurate and 17.8% did not correspond to a publication. Scenario variability had the greatest impact on ratings, followed by prompting strategy. Finally, latent knowledge across the fields showed minimal variation. In conclusion, ChatGPT provided useful responses in one-third of cases, but the number of errors and variability highlight that it is not yet adequate for everyday diagnostic practice and should be used with discretion as a support tool. The lack of thoroughness in providing references also suggests caution should be employed even when used as a self-learning tool. It is essential to recognize the irreplaceable role of human experts in synthesizing images, clinical data and experience for the intricate task of histopathological diagnosis.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"44 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140128449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Early diagnosis of transthyretin amyloidosis by detection of monomers in plasma microsamples using a protein crystal-based assay","authors":"Diogo Costa-Rodrigues, Leite Jose, Maria Joao Saraiva, Maria Rosario Almeida, Luis Gales","doi":"10.1101/2024.02.27.24303425","DOIUrl":"https://doi.org/10.1101/2024.02.27.24303425","url":null,"abstract":"Amyloid diseases are frequently associated with the appearance of an aberrant form of a protein, whose detection enables early diagnosis. In the case of transthyretin amyloidosis, the aberrant protein, the monomers, constitute the smallest species of the amyloid cascade, which creates engineering opportunities for sensing that remain virtually unexplored. Here, a two-step assay is devised, combining molecular sieving and immunodetection, for quantification of circulating monomeric transthyretin in the plasma. It is shown that mesoporous crystals built from biomolecules can selectively uptake transthyretin monomers up to measurable quantities. Furthermore, it was found that the use of endogenous molecules to produce the host framework drastically reduces unspecific adsorption of plasma proteins at the crystal surface, a feature that was observed with metal-organic frameworks. The assay was used to analyse plasma microsamples of patients and healthy controls. It shows a significant increase in the levels of monomeric transthyretin in the patients, proving its usefulness to establish the monomers as soluble and non-invasive marker of the disease. In addition, the assay can evaluate transthyretin stabilizers, an emergent strategy that broadened the treatment approach to the disease. Sensing the initial event of the transthyretin amyloid cascade with the proposed assay can make the difference for early diagnosis and eliminate the currently adopted invasive biopsies modalities for detection of the final products of the aggregation pathway.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140010353","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Public evidence on AI products for digital pathology","authors":"Gillian A Matthews, Clare McGenity, Daljeet Bansal, Darren Treanor","doi":"10.1101/2024.02.05.24302334","DOIUrl":"https://doi.org/10.1101/2024.02.05.24302334","url":null,"abstract":"Background: Novel products applying artificial intelligence (AI)-based approaches to digital pathology images have consistently emerged onto the commercial market, touting improvements in diagnostic accuracy, workflow efficiency, and treatment selection. However, publicly available information on these products can be variable, with few sources to obtain independent evidence.\u0000Methods: Our objective was to identify and assess the public evidence on AI-based products for digital pathology. We compared key features of products on the European Economic Area/Great Britain (EEA/GB) markets, including their regulatory approval, intended use, and published validation studies. We included products that used haematoxylin and eosin (H&E)-stained tissue images as input, applied an AI-based method to support image interpretation, and received regulatory approval by September 2023.\u0000Results: We identified 26 AI-based products that met our inclusion criteria. The majority (73%) were focused on breast pathology or uropathology, and their primary function was tumour or feature detection. Of the 26 products, 24 had received regulatory approval via the self-certification route as General in vitro diagnostic (IVD) medical devices, which does not require independent review by a conformity assessment body. Furthermore, only 10 of the products (38%) were associated with peer-reviewed scientific publications describing their development and internal validation, while 11 products (42%) had peer-reviewed publications describing external validation (i.e., testing on data from a source distinct to that used in development).\u0000Conclusions: The availability of public information on new products for digital pathology is struggling to keep up with the rapid pace of development. To support transparency, we gathered available public\u0000evidence on regulatory-approved AI products into an online register: https://resources.npic.uk/AI/ProductRegister. We anticipate this will provide an accessible resource on novel devices and support decisions on which products could bring benefit to patients.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"18 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139765070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Atypia of Undetermined Significance and ThyroSeq v3 Positive Call Rates as Quality Control Metrics for Cytology Laboratory Performance","authors":"Odille Mejia-Mejia, Andres Bravo-Gonzalez, Monica Sanchez-Avila, Youley Tjendra, Rodrigo Santoscoy, Katherine Drews-Elger, Yiqin Zuo, Camilo Arias-Abad, Carmen Gomez, Monica Garcia-Buitrago, Mehrdad Nadji, Merce Jorda, Jaylou M. Velez-Torres, Roberto Ruiz-Cordero","doi":"10.1101/2024.01.24.24301631","DOIUrl":"https://doi.org/10.1101/2024.01.24.24301631","url":null,"abstract":"<strong>Background</strong> The Bethesda system (TBS) for reporting thyroid cytopathology recommends an “atypia of undetermined significance (AUS)” rate of 10%. Recent data suggest that this category might be overused when the rate of cases with molecular positive results is low. As a quality metric, we calculated the AUS and positive call rates for our cytology lab and each cytopathologist.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"10 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139582142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-V-Stain: Multiplexed Virtual Staining of Histopathology Whole-Slide Images","authors":"Sonali Andani, Boqi Chen, Joanna Ficek-Pascual, Simon Heinke, Ruben Casanova, Bettina Sobottka, Bernd Bodenmiller, The Tumor Profiler Consortium, Viktor H Kölzer, Gunnar Rätsch","doi":"10.1101/2024.01.26.24301803","DOIUrl":"https://doi.org/10.1101/2024.01.26.24301803","url":null,"abstract":"Pathological assessment of Hematoxylin &amp; Eosin (H&amp;E) stained tissue samples is a well-established clinical routine for cancer diagnosis. While providing rich morphological data, it lacks information on protein expression patterns which is crucial for cancer prognosis and treatment recommendations. Imaging Mass Cytometry (IMC) excels in highly multiplexed protein profiling but faces challenges like high operational cost and a restrictive focus on small Regions-of-Interest. Addressing this, we introduce Multi-V-Stain, a novel image-to-image translation method for multiplexed IMC virtual staining. Our method effectively utilizes the rich morphological features from H&amp;E images to predict multiplexed protein expressions at a Whole-Slide Image level. In evaluations using an in-house melanoma dataset, Multi-V-Stain consistently outperforms existing methods in terms of image quality and biological relevance of the generated stains.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"327 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139582143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hemoglobin-Albumin-Lymphocyte-Platelet (HALP) score as a predictive value of Incidental prostate cancer for patients going for Transurethral resection of the prostate (TURP): A Single Center Study","authors":"Ahmed Bendari","doi":"10.1101/2024.01.23.24301670","DOIUrl":"https://doi.org/10.1101/2024.01.23.24301670","url":null,"abstract":"Aims: Prostate cancer is a significant health concern worldwide, and early detection is crucial for effective treatment. This study aimed to investigate the role of the Hemoglobin-Albumin-Lymphocyte-Platelet (HALP) score in detecting prostate cancer in patients undergoing Transurethral Resection of the Prostate (TURP). Additionally, comprehensive analysis was performed to explore clinical parameters associated with incidentally diagnosed prostate cancer post-TURP. Methods: A total of 131 patients with symptomatic bladder outlet obstruction who underwent TURP were included in the study. The patients were divided into two groups: those with benign prostatic hyperplasia (BPH) and those with invasive prostatic carcinoma. The IPC group consisted of patients with both low-grade and high-grade IPC determined by Gleason score. Demographic data, including age, race, medical history, body mass index, smoking and alcohol status, and family history of prostate cancer, were evaluated. Postoperative measurement of specimen weight and prostate-specific antigen (PSA) levels were also analyzed. Result: Results revealed that approximately 50% of patients had BPH, while the remaining 50% had IPC. Patients with IPC, particularly high-grade IPC, had significantly higher PSA levels and lower resected specimen weight compared to those with BPH. The HALP score, which incorporates hemoglobin, albumin, lymphocyte, and platelet levels, showed promise as a discriminatory tool for distinguishing between BPH and IPC, as well as between high-grade IPC and BPH/low-grade IPC. Logistic regression analysis identified increased PSA levels, decreased HALP score, and smaller specimen weight as independent predictive factors for IPC after TURP. Notably, HALP score was the only significant independent predictive factor associated with high-grade IPC. Conclusion: These findings contribute to the understanding of risk factors and diagnostic tools for incidentally detected prostate cancer in patients with bladder outlet obstruction undergoing TURP. The HALP score, along with PSA levels and specimen weight, can aid in the early detection and management of prostate cancer. Further research is warranted to validate these findings and explore the clinical utility of the HALP score in predicting prostate cancer outcomes.","PeriodicalId":501528,"journal":{"name":"medRxiv - Pathology","volume":"53 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139553930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}