Journal of Pathology Informatics最新文献

Treatment of patients with oesophageal and gastric cancer (OeGC) is guided by disease stage, patient performance status and preferences. Lymph node (LN) status is one of the strongest prognostic factors for OeGC patients. However, survival varies between patients with the same disease stage and LN status. We recently showed that LN size from patients with OeGC might also have prognostic value, thus making delineations of LNs essential for size estimation and the extraction of other imaging biomarkers.

We hypothesized that a machine learning workflow is able to: (1) find digital H&E stained slides containing LNs, (2) create a scoring system providing degrees of certainty for the results, and (3) delineate LNs in those images.

To train and validate the pipeline, we used 1695 H&E slides from the OE02 trial. The dataset was divided into training (80%) and validation (20%). The model was tested on an external dataset of 826 H&E slides from the OE05 trial. U-Net architecture was used to generate prediction maps from which predefined features were extracted. These features were subsequently used to train an XGBoost model to determine if a region truly contained a LN. With our innovative method, the balanced accuracies of the LN detection were 0.93 on the validation dataset (0.83 on the test dataset) compared to 0.81 (0.81) on the validation (test) datasets when using the standard method of thresholding U-Net predictions to arrive at a binary mask. Our method allowed for the creation of an “uncertain” category, and partly limited false-positive predictions on the external dataset. The mean Dice score was 0.73 (0.60) per-image and 0.66 (0.48) per-LN for the validation (test) datasets.

Our pipeline detects images with LNs more accurately than conventional methods, and high-throughput delineation of LNs can facilitate future LN content analyses of large datasets.

In medical image diagnosis, identifying the attention region, i.e., the region of interest for which the diagnosis is made, is an important task. Various methods have been developed to automatically identify target regions from given medical images. However, in actual medical practice, the diagnosis is made based on both the images and various clinical records. Consequently, pathologists examine medical images with prior knowledge of the patients and the attention regions may change depending on the clinical records. In this study, we propose a method, called the Personalized Attention Mechanism (PersAM) method, by which the attention regions in medical images according to the clinical records. The primary idea underlying the PersAM method is the encoding of the relationships between medical images and clinical records using a variant of the Transformer architecture. To demonstrate the effectiveness of the PersAM method, we applied it to a large-scale digital pathology problem involving identifying the subtypes of 842 malignant lymphoma patients based on their gigapixel whole-slide images and clinical records.

Artificial Intelligence is a tool poised to transform healthcare, with use in diagnostics and therapeutics. The widespread use of digital pathology has been due to the advent of whole slide imaging. Cheaper storage for digital images, along with unprecedented progress in artificial intelligence, have paved the synergy of these two fields. This has pushed the limits of traditional diagnosis using light microscopy, from a more subjective to a more objective method of looking at cases, incorporating grading too. The grading of histopathological images of urothelial carcinoma of the urinary bladder is important with direct implications for surgical management and prognosis.

In this study, the aim is to classify urothelial carcinoma into low and high grade based on the WHO 2016 classification. The hematoxylin and eosin-stained transurethral resection of bladder tumor (TURBT) samples of both low and high grade non-invasive papillary urothelial carcinoma were digitally scanned. Patches were extracted from these whole slide images to feed into a deep learning (Convolution Neural Network: CNN) model. Patches were segregated if they had tumor tissue and only included for model training if a threshold of 90% of tumor tissue per patch was seen. Various parameters of the deep learning model, known as hyperparameters, were optimized to get the best accuracy for grading or classification into low- and high-grade urothelial carcinoma. The model was robust with an overall accuracy of 90% after hyperparameter tuning. Visualization in the form of a class activation map using Grad-CAM was done. This indicates that such a model can be used as a companion diagnostic tool for grading of urothelial carcinoma. The probable causes of this accuracy are summarized along with the limitations of this study and future work possible.

Disease interpretation by computer-aided diagnosis systems in digital pathology depends on reliable detection and segmentation of nuclei in hematoxylin and eosin (HE) images. These 2 tasks are challenging since appearance of both cell nuclei and background structures are very variable. This paper presents a method to improve nuclei detection and segmentation in HE images by removing tiles that only contain background information. The method divides each image into smaller patches and uses their projection to the noiselet space to capture different spatial features from non-nuclei background and nuclei structures. The noiselet features are clustered by a K-means algorithm and the resultant partition, defined by the cluster centroids, is herein named the noiselet code-book. A part of an image, a tile, is divided into patches and represented by the histogram of occurrences of the projected patches in the noiselet code-book. Finally, with these histograms, a classifier learns to differentiate between nuclei and non-nuclei tiles. By applying a conventional watershed-marked method to detect and segment nuclei, evaluation consisted in comparing pure watershed method against denoising-plus-watershed in an open database with 8 different types of tissues. The averaged F-score of nuclei detection improved from 0.830 to 0.86 and the dice score after segmentation increased from 0.701 to 0.723.

Deep learning has been effective for histology image analysis in digital pathology. However, many current deep learning approaches require large, strongly- or weakly labeled images and regions of interest, which can be time-consuming and resource-intensive to obtain. To address this challenge, we present HistoPerm, a view generation method for representation learning using joint embedding architectures that enhances representation learning for histology images. HistoPerm permutes augmented views of patches extracted from whole-slide histology images to improve classification performance. We evaluated the effectiveness of HistoPerm on 2 histology image datasets for Celiac disease and Renal Cell Carcinoma, using 3 widely used joint embedding architecture-based representation learning methods: BYOL, SimCLR, and VICReg. Our results show that HistoPerm consistently improves patch- and slide-level classification performance in terms of accuracy, F1-score, and AUC. Specifically, for patch-level classification accuracy on the Celiac disease dataset, HistoPerm boosts BYOL and VICReg by 8% and SimCLR by 3%. On the Renal Cell Carcinoma dataset, patch-level classification accuracy is increased by 2% for BYOL and VICReg, and by 1% for SimCLR. In addition, on the Celiac disease dataset, models with HistoPerm outperform the fully supervised baseline model by 6%, 5%, and 2% for BYOL, SimCLR, and VICReg, respectively. For the Renal Cell Carcinoma dataset, HistoPerm lowers the classification accuracy gap for the models up to 10% relative to the fully supervised baseline. These findings suggest that HistoPerm can be a valuable tool for improving representation learning of histopathology features when access to labeled data is limited and can lead to whole-slide classification results that are comparable to or superior to fully supervised methods.

Patient portals allow patients to access their personal health information. The 21st Century Cures Act in the United States sought to eliminate ‘information blocking’, requiring timely release upon request of electronic health information including diagnostic test results. Some health systems, including the one in the present study, chose a systematic switch to immediate release of all or nearly all diagnostic test results to patient portals as part of compliance with the Cures Act. Our primary objective was to study changes in the time to view test results by patients before and after implementation of Cures Act-related changes. This retrospective pre-post study included data from two 10-month time periods before and after implementation of Cures Act-related changes at an academic medical center. The study included all patients (adult and pediatric) with diagnostic testing (laboratory and imaging) performed in the outpatient, inpatient, or emergency department settings. Between February 9, 2020 and December 9, 2021, there was a total of 3 809 397 diagnostic tests from 204 605 unique patients (3 320 423 tests for adult patients; 488 974 for pediatric patients). Overall, 56.5% (115 627) of patients were female, 84.1% (172 048) white, and 96.5% (197 517) preferred English as primary language. The odds of viewing test results within 1 and 30 days after portal release increased monthly throughout both time periods before and after the Cures Act for all patients. The rate of increase was significantly higher after implementation only in the subgroup of tests belonging to adult patients with active MyChart accounts. Immediate release shifted a higher proportion of result/report release to weekends (3.2% pre-Cures vs 15.3% post-Cures), although patient viewing patterns by day of week and time of day were similar before and after immediate release changes. The switch to immediate release of diagnostic test results to the patient portal resulted in a higher fraction of results viewed within 1 day across outpatient, inpatient, and emergency department settings.

Pathologists need to compare histopathological images of normal and diseased tissues between different samples, cases, and species. We have designed an interactive system, termed Comparative Pathology Workbench (CPW), which allows direct and dynamic comparison of images at a variety of magnifications, selected regions of interest, as well as the results of image analysis or other data analyses such as scRNA-seq. This allows pathologists to indicate key diagnostic features, with a mechanism to allow discussion threads amongst expert groups of pathologists and other disciplines. The data and associated discussions can be accessed online from anywhere in the world. The Comparative Pathology Workbench (CPW) is a web-browser-based visual analytics platform providing shared access to an interactive “spreadsheet” style presentation of image and associated analysis data. The CPW provides a grid layout of rows and columns so that images that correspond to matching data can be organised in the form of an image-enabled “spreadsheet”. An individual workbench can be shared with other users with read-only or full edit access as required. In addition, each workbench element or the whole bench itself has an associated discussion thread to allow collaborative analysis and consensual interpretation of the data.

The CPW is a Django-based web-application that hosts the workbench data, manages users, and user-preferences. All image data are hosted by other resource applications such as OMERO or the Digital Slide Archive. Further resources can be added as required. The discussion threads are managed using WordPress and include additional graphical and image data. The CPW has been developed to allow integration of image analysis outputs from systems such as QuPath or ImageJ. All software is open-source and available from a GitHub repository.

Training models with semi- or self-supervised learning methods is one way to reduce annotation effort since they rely on unlabeled or sparsely labeled datasets. Such approaches are particularly promising for domains with a time-consuming annotation process requiring specialized expertise and where high-quality labeled machine learning datasets are scarce, like in computational pathology. Even though some of these methods have been used in the histopathological domain, there is, so far, no comprehensive study comparing different approaches. Therefore, this work compares feature extractors models trained with state-of-the-art semi- or self-supervised learning methods PAWS, SimCLR, and SimSiam within a unified framework. We show that such models, across different architectures and network configurations, have a positive performance impact on histopathological classification tasks, even in low data regimes. Moreover, our observations suggest that features learned from a particular dataset, i.e., tissue type, are only in-domain transferable to a certain extent. Finally, we share our experience using each method in computational pathology and provide recommendations for its use.

Introduction

The transition to digital pathology has been carried out by several laboratories across the globe, with some cases described in Portugal. In this article, we describe the transition to digital pathology in a high-volume private laboratory, considering the main challenges and opportunities.

Material and methods

Our process started in 2020, with laboratory workflow adaptation and we are currently using a high-capacity scanner (Aperio GT450DX) to digitize slides at 20×. The visualization system, Aperio eSlide Manager WebViewer, is integrated into the Laboratory System. The validation process followed the Royal College of Pathologists Guidelines.

Results

Regarding validation, the first phase detected an error rate of 6.8%, mostly due to digitization errors. Phase optimization and collaboration with technical services led to improvements in this process. In the second validation phase, most of the slides had the desired quality for evaluation, with only an error rate of 0.6%, corrected with a new scan. The interpathologist correlation had a total agreement rate of 96.87% and 3.13% partial agreement.

Conclusion

The implementation and validation of digital pathology was a success, being ready for prime time. The total integration of all laboratory systems and the acquisition of new equipment will maximize their use, especially with the application of artificial intelligence algorithms.

While VCF formatted files are the lingua franca of next-generation sequencing, most EHRs do not provide native VCF support. As a result, labs often must send non-structured PDF reports to the EHR. On the other hand, while FHIR adoption is growing, most EHRs support HL7 interoperability standards, particularly those based on the HL7 Version 2 (HL7v2) standard. The HL7 Version 2 genomics component of the HL7 Laboratory Results Interface (HL7v2 LRI) standard specifies a formalism for the structured communication of genomic data from lab to EHR. We previously described an open-source tool (vcf2fhir) that converts VCF files into HL7 FHIR format. In this report, we describe how the utility has been extended to output HL7v2 LRI data that contains both variants and variant annotations (e.g., predicted phenotypes and therapeutic implications). Using this HL7v2 converter, we implemented an automated pipeline for moving structured genomic data from the clinical laboratory to EHR. We developed an open source hl7v2GenomicsExtractor that converts genomic interpretation report files into a series of HL7v2 observations conformant to HL7v2 LRI. We further enhanced the converter to produce output conformant to Epic's genomic import specification and to support alternative input formats. An automated pipeline for pushing standards-based structured genomic data directly into the EHR was successfully implemented, where genetic variant data and the clinical annotations are now both available to be viewed in the EHR through Epic's genomics module. Issues encountered in the development and deployment of the HL7v2 converter primarily revolved around data variability issues, primarily lack of a standardized representation of data elements within various genomic interpretation report files. The technical implementation of a HL7v2 message transformation to feed genomic variant and clinical annotation data into an EHR has been successful. In addition to genetic variant data, the implementation described here releases the valuable asset of clinically relevant genomic annotations provided by labs from static PDFs to calculable, structured data in EHR systems.