-
Laboratory Investigation; a Journal of... Jul 2019Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and...
Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and overlapping morphological phenotypes that obscure classification and quantitative assessments. Recently, powerful machine learning-based approaches have emerged, allowing the recognition and quantification of pathological changes from digital images. Here, we applied deep learning to the neuropathological assessment of NFT in postmortem human brain tissue to develop a classifier capable of recognizing and quantifying tau burden. The histopathological material was derived from 22 autopsy brains from patients with tauopathies. We used a custom web-based informatics platform integrated with an in-house information management system to manage whole slide images (WSI) and human expert annotations as ground truth. We utilized fully annotated regions to train a deep learning fully convolutional neural network (FCN) implemented in PyTorch against the human expert annotations. We found that the deep learning framework is capable of identifying and quantifying NFT with a range of staining intensities and diverse morphologies. With our FCN model, we achieved high precision and recall in naive WSI semantic segmentation, correctly identifying tangle objects using a SegNet model trained for 200 epochs. Our FCN is efficient and well suited for the practical application of WSIs with average processing times of 45 min per WSI per GPU, enabling reliable and reproducible large-scale detection of tangles. We measured performance on test data of 50 pre-annotated regions on eight naive WSI across various tauopathies, resulting in the recall, precision, and an F1 score of 0.92, 0.72, and 0.81, respectively. Machine learning is a useful tool for complex pathological assessment of AD and other tauopathies. Using deep learning classifiers, we have the potential to integrate cell- and region-specific annotations with clinical, genetic, and molecular data, providing unbiased data for clinicopathological correlations that will enhance our knowledge of the neurodegeneration.
Topics: Aged; Aged, 80 and over; Brain; Deep Learning; Female; Humans; Male; Neuropathology; Tauopathies
PubMed: 30770886
DOI: 10.1038/s41374-019-0202-4 -
PloS One 2021Pathologists generally pan, focus, zoom and scan tissue biopsies either under microscopes or on digital images for diagnosis. With the rapid development of whole-slide...
Pathologists generally pan, focus, zoom and scan tissue biopsies either under microscopes or on digital images for diagnosis. With the rapid development of whole-slide digital scanners for histopathology, computer-assisted digital pathology image analysis has attracted increasing clinical attention. Thus, the working style of pathologists is also beginning to change. Computer-assisted image analysis systems have been developed to help pathologists perform basic examinations. This paper presents a novel lightweight detection framework for automatic tumor detection in whole-slide histopathology images. We develop the Double Magnification Combination (DMC) classifier, which is a modified DenseNet-40 to make patch-level predictions with only 0.3 million parameters. To improve the detection performance of multiple instances, we propose an improved adaptive sampling method with superpixel segmentation and introduce a new heuristic factor, local sampling density, as the convergence condition of iterations. In postprocessing, we use a CNN model with 4 convolutional layers to regulate the patch-level predictions based on the predictions of adjacent sampling points and use linear interpolation to generate a tumor probability heatmap. The entire framework was trained and validated using the dataset from the Camelyon16 Grand Challenge and Hubei Cancer Hospital. In our experiments, the average AUC was 0.95 in the test set for pixel-level detection.
Topics: Humans; Image Processing, Computer-Assisted; Neoplasm Metastasis; Pathology
PubMed: 33979398
DOI: 10.1371/journal.pone.0251521 -
Cellular Oncology : the Official... 2007Since 1991, a nationwide histopathology and cytopathology network and archive is in operation in The Netherlands under the name PALGA, encompassing all sixty-four...
Since 1991, a nationwide histopathology and cytopathology network and archive is in operation in The Netherlands under the name PALGA, encompassing all sixty-four pathology laboratories in The Netherlands. The overall system comprises decentralized systems at the participating laboratories, a central databank, and a dedicated communication and information exchange tool. Excerpts of all histopathology and cytopathology reports are generated automatically at the participating laboratories and transferred to the central databank. Both the decentralized systems and the central system perform checks on the quality and completeness of excerpts. Currently, about 42 million records on almost 10 million patients are stored in the central databank. Each excerpt contains patient identifiers, including demographic data and the so-called PALGA diagnosis. The latter is structured along five classification axes: topography, morphology, function, procedure, and diseases. All data transfer and communication occurs electronically with encryption of patient and laboratory identifiers. All excerpts are continuously available to all participating pathology laboratories, thus contributing to the quality of daily patient care. In addition, external parties may obtain permission to use data from the PALGA system, either on an ongoing basis or on the basis of a specific permission. Annually, 40 to 60 applications for permission to use PALGA data are submitted. Among external users are the Dutch cancer registry, population-based screening programs for cancer of the uterine cervix and breast cancer in The Netherlands, and individual investigators addressing a range of research questions. Many scientific papers and theses incorporating PALGA data have been published already. In conclusion, the PALGA system is a unique system that requires a minimal effort on the part of the participating laboratories, while providing them a powerful tool in their daily practices.
Topics: Biological Specimen Banks; Humans; Information Systems; National Health Programs; Netherlands; Pathology, Clinical
PubMed: 17429138
DOI: 10.1155/2007/971816 -
Archives of Pathology & Laboratory... Aug 2003Laboratories contemplating either the addition of new molecular tests or modifying methods approved by the Food and Drug Administration for human papillomavirus testing... (Review)
Review
Laboratories contemplating either the addition of new molecular tests or modifying methods approved by the Food and Drug Administration for human papillomavirus testing should be aware of a variety of procedural, performance, and regulatory issues surrounding such activity. Diagnostic medical laboratory testing in the United States is regulated by the Centers for Medicare and Medicaid Services, an agency formerly known as the Health Care Finance Administration. The regulatory vehicle of the Centers for Medicare and Medicaid Services is manifested in the Clinical Laboratory Improvement Amendments (CLIA). The CLIA program has put into place specific regulations for laboratory quality control, which includes specific recommendations for method validation. Regulations that must be followed regarding personnel, quality control, quality assurance, method validation, and proficiency testing depend on the complexity category of the individual test. All molecular diagnostic tests, including those for human papillomavirus, are considered high complexity. The Centers for Medicare and Medicaid Services retains the authority to allow private, national accreditation organizations to "deem" that a laboratory is compliant with CLIA '88 requirements. Accreditation organizations, such as the Joint Commission for Accreditation of Hospitals, the Commission on Office Laboratory Accreditation, and the College of American Pathologists (CAP), as well as several state medical laboratory-accrediting agencies, possess the authority to deem laboratories as "CLIA-approved." The CAP, through its Laboratory Accreditation Program, has promoted standards for laboratory performance and method validation. In general, guidelines set forth in the CAP Laboratory Accreditation Program checklists specify that all clinical laboratory testing must essentially meet those requirements defined for high-complexity testing under CLIA '88, including test validation standards, reportable/reference ranges, performance criteria, and proficiency testing.
Topics: Female; Humans; Molecular Diagnostic Techniques; Papillomaviridae; Papillomavirus Infections; Pathology, Clinical; Practice Guidelines as Topic; Quality Assurance, Health Care; Reproducibility of Results; Tumor Virus Infections; Uterine Cervical Neoplasms
PubMed: 12873170
DOI: 10.5858/2003-127-978-AQRRAM -
AMA Journal of Ethics Aug 2016The 2015 Institute of Medicine report on diagnostic error has placed a national spotlight on the importance of improving communication among clinicians and between...
The 2015 Institute of Medicine report on diagnostic error has placed a national spotlight on the importance of improving communication among clinicians and between clinicians and patients [1]. The report emphasizes the critical role that communication plays in patient safety and outlines ways that pathologists can support this process. Despite recognition of communication as an essential element in patient care, pathologists currently undergo limited (if any) formal training in communication skills. To address this gap, we at the University of Washington Medical Center developed communication training with the goal of establishing best practice procedures for effective pathology communication. The course includes lectures, role playing, and simulated clinician-pathologist interactions for training and evaluation of pathology communication performance. Providing communication training can help create reliable communication pathways that anticipate and address potential barriers and errors before they happen.
Topics: Communication; Curriculum; Humans; Interprofessional Relations; Pathologists; Pathology; Physician-Patient Relations; Schools, Medical; Washington
PubMed: 27550564
DOI: 10.1001/journalofethics.2016.18.8.medu1-1608 -
American Journal of Clinical Pathology Apr 2021The ongoing global severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic necessitates adaptations in the practice of surgical pathology at scale. Primary...
OBJECTIVES
The ongoing global severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic necessitates adaptations in the practice of surgical pathology at scale. Primary diagnosis by whole-slide imaging (WSI) is a key component that would aid departments in providing uninterrupted histopathology diagnosis and maintaining revenue streams from disruption. We sought to perform rapid validation of the use of WSI in primary diagnosis meeting recommendations of the College of American Pathologists guidelines.
METHODS
Glass slides from clinically reported cases from 5 participating pathologists with a preset washout period were digitally scanned and reviewed in settings identical to typical reporting. Cases were classified as concordant or with minor or major disagreement with the original diagnosis. Randomized subsampling was performed, and mean concordance rates were calculated.
RESULTS
In total, 171 cases were included and distributed equally among participants. For the group as a whole, the mean concordance rate in sampled cases (n = 90) was 83.6% counting all discrepancies and 94.6% counting only major disagreements. The mean pathologist concordance rate in sampled cases (n = 18) ranged from 90.49% to 97%.
CONCLUSIONS
We describe a novel double-blinded method for rapid validation of WSI for primary diagnosis. Our findings highlight the occurrence of a range of diagnostic reproducibility when deploying digital methods.
Topics: COVID-19; Double-Blind Method; Humans; Image Interpretation, Computer-Assisted; Observer Variation; Pandemics; Pathology, Surgical; Practice Guidelines as Topic; Reproducibility of Results; Retrospective Studies; Telepathology
PubMed: 33511392
DOI: 10.1093/ajcp/aqaa280 -
International Journal of Legal Medicine May 2021The motor vehicle crash (MVC) constitutes an important challenge for forensic pathology in order to identify the manner and cause of death. Our study focuses on a fatal...
INTRODUCTION
The motor vehicle crash (MVC) constitutes an important challenge for forensic pathology in order to identify the manner and cause of death. Our study focuses on a fatal accident during a rally race corresponding to MVC sub-category.
MATERIALS AND METHOD
Postmortem computed tomography (PMCT) was performed before the conventional autopsy. Autoptic and PMCT data were compared. Data collection allowed analyzing biomechanical dynamics of the incident and post-traumatic injuries through qualitative-statistics and solicitation quantitative indices.
RESULTS
Photo and circumstantial evidence analysis showed a wrong installation of double shoulder belt system of head and neck support (HANS) collar. PMTC clearly highlighted multiple and bilateral fractures involving roof and base of skull; a displaced fracture of the right acetabulum was also encountered. Autopsy confirmed PMCT data and revealed a brainstem laceration. AIS (Abbreviated Injury Scale) achieved a maximum score in consideration of fatal injuries.
DISCUSSION
The injuries analysis resulting from photographic surveys examination, conventional autopsy, and PMCT has led us to confirm a fatal front collision with a tree trunk. Head trauma represents a major injury in the present case. In this case, head injuries, related to whiplash trauma, are a consequence of a double shoulder belt system (HANS collar component) wrong installation.
CONCLUSION
MVC and especially high-speed motor racing represent an important death cause. There was, for this reason, a marked development of cars and occupants' safety systems, such as HANS collar. PMCT improves the diagnostic performance of conventional autopsy and increases forensic medical knowledge related to traumatic injuries.
Topics: Abbreviated Injury Scale; Athletic Injuries; Autopsy; Biomechanical Phenomena; Craniocerebral Trauma; Fatal Outcome; Forensic Pathology; Fractures, Bone; Humans; Male; Motor Vehicles; Protective Devices; Tomography, X-Ray Computed
PubMed: 33237457
DOI: 10.1007/s00414-020-02470-2 -
JCO Clinical Cancer Informatics Apr 2019Digital pathology (DP), referring to the digitization of tissue slides, is beginning to change the landscape of clinical diagnostic workflows and has engendered active...
PURPOSE
Digital pathology (DP), referring to the digitization of tissue slides, is beginning to change the landscape of clinical diagnostic workflows and has engendered active research within the area of computational pathology. One of the challenges in DP is the presence of artefacts and batch effects, unintentionally introduced during both routine slide preparation (eg, staining, tissue folding) and digitization (eg, blurriness, variations in contrast and hue). Manual review of glass and digital slides is laborious, qualitative, and subject to intra- and inter-reader variability. Therefore, there is a critical need for a reproducible automated approach of precisely localizing artefacts to identify slides that need to be reproduced or regions that should be avoided during computational analysis.
METHODS
Here we present HistoQC, a tool for rapidly performing quality control to not only identify and delineate artefacts but also discover cohort-level outliers (eg, slides stained darker or lighter than others in the cohort). This open-source tool employs a combination of image metrics (eg, color histograms, brightness, contrast), features (eg, edge detectors), and supervised classifiers (eg, pen detection) to identify artefact-free regions on digitized slides. These regions and metrics are presented to the user via an interactive graphical user interface, facilitating artefact detection through real-time visualization and filtering. These same metrics afford users the opportunity to explicitly define acceptable tolerances for their workflows.
RESULTS
The output of HistoQC on 450 slides from The Cancer Genome Atlas was reviewed by two pathologists and found to be suitable for computational analysis more than 95% of the time.
CONCLUSION
These results suggest that HistoQC could provide an automated, quantifiable, quality control process for identifying artefacts and measuring slide quality, in turn helping to improve both the repeatability and robustness of DP workflows.
Topics: Humans; Neoplasms; Pathology, Molecular; Quality Control; Reproducibility of Results; Software; User-Computer Interface; Workflow
PubMed: 30990737
DOI: 10.1200/CCI.18.00157 -
Archives of Pathology & Laboratory... Mar 2014In the late 1990s, the Accreditation Council for Graduate Medical Education developed the Outcomes Project and the 6 general competencies with the intent to improve the...
CONTEXT
In the late 1990s, the Accreditation Council for Graduate Medical Education developed the Outcomes Project and the 6 general competencies with the intent to improve the outcome of graduate medical education in the United States. The competencies were used as the basis for developing learning goals and objectives and tools to evaluate residents' performance. By the mid-2000s the stakeholders in resident education and the general public felt that the Outcomes Project had fallen short of expectations.
OBJECTIVE
To develop a new evaluation method to track trainee progress throughout residency using benchmarks called milestones. A change in leadership at the Accreditation Council for Graduate Medical Education brought a new vision for the accreditation of training programs and a radically different approach to the evaluation of residents.
DATA SOURCES
The Pathology Milestones Working Group reviewed examples of developing milestones in other specialties, the literature, and the Accreditation Council for Graduate Medical Education program requirements for pathology to develop pathology milestones. The pathology milestones are a set of objective descriptors for measuring progress in the development of competency in patient care, procedural skill sets, medical knowledge, practice-based learning and improvement, interpersonal and communication skills, professionalism, and systems-based practice.
CONCLUSIONS
The milestones provide a national standard for evaluation that will be used for the assessment of all residents in Accreditation Council for Graduate Medical Education-accredited pathology training programs.
Topics: Accreditation; Clinical Competence; Education, Medical, Graduate; Humans; Pathology; United States
PubMed: 24576024
DOI: 10.5858/arpa.2013-0260-SA -
Turk Patoloji Dergisi 2019Intraoperative consultations or frozen sections for central nervous system (CNS) tumors present a significant challenge for surgical pathologists because of their... (Review)
Review
Intraoperative consultations or frozen sections for central nervous system (CNS) tumors present a significant challenge for surgical pathologists because of their relative rarity and diversity. Yet, such lesions are encountered by every surgical pathologist, and a basic understanding of clinical, radiological and genetic information is critical to successfully evaluate CNS frozen sections. It is often beneficial to have a systematic approach or an algorithm, and to be aware of the common pitfalls and mimickers when dealing with these lesions. We propose such an algorithm in an effort to construct a sensible approach to CNS frozen sections that considers recent developments in the WHO CNS tumor classification. The algorithm was developed for surgical pathologists who are occasionally faced with making diagnosis of CNS tumors on frozen sections. To test the algorithm and its practicability, we selected a group of tumors among a total of 3288 consecutive intraoperative consultations performed at UCSF between 2013 and 2017. The selected cases represented lesions that may be encountered in everyday surgical pathology and constituted a fair reflection of the main group. The algorithm was used by three of the authors who did not have formal neuropathology training and had been in surgical pathology practice for at least 3 years. There was a very high level of concordance among the authors' diagnosis (interobserver concordance: 0.83-0.97-kappa value) using the algorithm with high intraobserver reliability (concordance 93%, p < 0.001). We suggest that an algorithmic approach is an effective means for the surgical pathologists, and may help reach diagnosis during frozen sections.
Topics: Algorithms; Central Nervous System Neoplasms; Frozen Sections; Humans; Pathology, Surgical
PubMed: 31107540
DOI: 10.5146/tjpath.2018.01460