Journal of Pathology Informatics最新文献

The success of immuno-oncology treatments promises long-term cancer remission for an increasing number of patients. The response to checkpoint inhibitor drugs has shown a correlation with the presence of immune cells in the tumor and tumor microenvironment. An in-depth understanding of the spatial localization of immune cells is therefore critical for understanding the tumor’s immune landscape and predicting drug response. Computer-aided systems are well suited for efficiently quantifying immune cells in their spatial context. Conventional image analysis approaches are often based on color features and therefore require a high level of manual interaction. More robust image analysis methods based on deep learning are expected to decrease this reliance on human interaction and improve the reproducibility of immune cell scoring. However, these methods require sufficient training data and previous work has reported low robustness of these algorithms when they are tested on out-of-distribution data from different pathology labs or samples from different organs. In this work, we used a new image analysis pipeline to explicitly evaluate the robustness of marker-labeled lymphocyte quantification algorithms depending on the number of training samples before and after being transferred to a new tumor indication. For these experiments, we adapted the RetinaNet architecture for the task of T-lymphocyte detection and employed transfer learning to bridge the domain gap between tumor indications and reduce the annotation costs for unseen domains. On our test set, we achieved human-level performance for almost all tumor indications with an average precision of 0.74 in-domain and 0.72–0.74 cross-domain. From our results, we derive recommendations for model development regarding annotation extent, training sample selection, and label extraction for the development of robust algorithms for immune cell scoring. By extending the task of marker-labeled lymphocyte quantification to a multi-class detection task, the pre-requisite for subsequent analyses, e.g., distinguishing lymphocytes in the tumor stroma from tumor-infiltrating lymphocytes, is met.

The development of rapid and accurate Whole Slide Imaging (WSI) has paved the way for the application of Artificial Intelligence (AI) to digital pathology. The availability of WSI in the recent years allowed the rapid development of various AI technologies to blossom. WSI-based digital pathology combined with neural networks can automate arduous and time-consuming tasks of slide evaluation. Machine Learning (ML)-based AI has been demonstrated to outperform pathologists by eliminating inter- and intra-observer subjectivity, obtaining quantitative data from slide images, and extracting hidden image patterns that are relevant to disease subtype and progression. In this review, we outline the functionality of different AI technologies such as neural networks and deep learning and discover how aspects of different diseases make them benefit from the implementation of AI. AI has proven to be valuable in many different organs, with this review focusing on the liver, kidney, and lungs. We also discuss how AI and image analysis not only can grade diseases objectively but also discover aspects of diseases that have prognostic value. In the end, we review the current status of the integration of AI in pathology and share our vision on the future of digital pathology.

Background

Large bowel biopsies are one of the commonest types of biopsy specimen. We describe a service evaluation study to test the feasibility of using artificial intelligence (AI) to triage large bowel biopsies from a reporting backlog and prioritize those that require more urgent reporting.

Methods

The pathway was developed in the UK by National Health Service (NHS) laboratory staff working in a medium-sized general hospital.   The AI platform was interfaced with the slide scanner software and the reporting platform’s software, so that pathologists could correct the AI label and reinforce the training set as they reported the cases.

Results

The AI classifier achieved a sensitivity of 97.56% and specificity of 93.02% for the case-level-diagnosis of neoplasia (adenoma and adenocarcinoma) and for an AI diagnosis of any significant pathology (i.e., adenomas, adenocarcinomas, inflammation, hyperplastic polyps, and sessile serrated lesions) sensitivity was 95.65% and specificity 92.96%. The automated AI diagnostic classification pathway took approximately 175 s per slide to download and process the scanned whole slide image (WSI) and return an AI diagnostic classification. Biopsies with an AI diagnosis of neoplasia or inflammation were prioritized for reporting while the remainder followed the routine reporting pathway. The AI triaged pathway resulted in a significantly shorter reporting turnaround time for pathologist verified neoplastic cases (P < 0.001) and inflammation (P < 0.05). The project’s costs amounted to  £14800, excluding laboratory staff salaries. More time and resources were spent on developing the interface between the AI platform and laboratory IT systems than on the development of the AI platform itself.

Conclusions

NHS laboratory staff were able to implement an AI solution to accurately triage large bowel biopsies into several diagnostic classes and this improved reporting turnaround times for cases with neoplasia or with inflammation.

Objectives

This study aimed to validate a digital image analysis (DIA) workflow for automatic positive cell detection and positive region delineation for immunohistochemical hypoxia markers with a nuclear (hypoxia-inducible factor 1α [HIF-1α]) and a cytoplasmic (pimonidazole [PIMO]) staining pattern.

Materials and methods

101 tissue fragments from 44 laryngeal tumor biopsies were immunohistochemically stained for HIF-1α and PIMO. QuPath was used to determine the percentage of positive cells and to delineate positive regions automatically. For HIF-1α, only cells with strong staining were considered positive. Three dedicated head and neck pathologists scored the percentage of positive cells using three categories (0: <1%; 1: 1%–33%; 2: >33%;). The pathologists also delineated the positive regions on 14 corresponding PIMO and HIF-1α-stained fragments. The consensus between observers was used as the reference standard and was compared to the automatic delineation.

Results

Agreement between categorical positivity scores was 76.2% and 65.4% for PIMO and HIF-1α, respectively. In all cases of disagreement in HIF-1α fragments, the DIA underestimated the percentage of positive cells. As for the region detection, the DIA correctly detected most positive regions on PIMO fragments (false positive area=3.1%, false negative area=0.7%). In HIF-1α, the DIA missed some positive regions (false positive area=1.3%, false negative area=9.7%).

Conclusions

Positive cell and region detection on biopsy material is feasible, but further optimization is needed before unsupervised use. Validation at varying DAB staining intensities is hampered by lack of reliability of the gold standard (i.e., visual human interpretation). Nevertheless, the DIA method has the potential to be used as a tool to assist pathologists in the analysis of IHC staining.

Digital pathology technologies, including whole slide imaging (WSI), have significantly improved modern clinical practices by facilitating storing, viewing, processing, and sharing digital scans of tissue glass slides. Researchers have proposed various artificial intelligence (AI) solutions for digital pathology applications, such as automated image analysis, to extract diagnostic information from WSI for improving pathology productivity, accuracy, and reproducibility. Feature extraction methods play a crucial role in transforming raw image data into meaningful representations for analysis, facilitating the characterization of tissue structures, cellular properties, and pathological patterns. These features have diverse applications in several digital pathology applications, such as cancer prognosis and diagnosis. Deep learning-based feature extraction methods have emerged as a promising approach to accurately represent WSI contents and have demonstrated superior performance in histology-related tasks. In this survey, we provide a comprehensive overview of feature extraction methods, including both manual and deep learning-based techniques, for the analysis of WSIs. We review relevant literature, analyze the discriminative and geometric features of WSIs (i.e., features suited to support the diagnostic process and extracted by “engineered” methods as opposed to AI), and explore predictive modeling techniques using AI and deep learning. This survey examines the advances, challenges, and opportunities in this rapidly evolving field, emphasizing the potential for accurate diagnosis, prognosis, and decision-making in digital pathology.

Background

GP63, also known as Leishmanolysin, is a multifunctional virulence factor abundant on the surface of Leishmania spp. small peptides with anticancer capabilities that are selective and toxic to cancer cells are known as anticancer peptides. We aimed to demonstrate the activity of GP63 and its anticancer properties on melanoma using a range of in silico tools and screening methods to identify predicted and designed anticancer peptides.

Methods

Various in silico modeling methodologies are used to establish the three-dimensional (3D) structure of GP63. Refinement and re-evaluation of the modeled structures and the built models' quality evaluated using the different docking used to find the interacting amino acids between MMP2 and GP63 and its anticancer peptides. AntiCP2.0 is used for screening anticancer peptides. 2D interaction plots of protein–ligand complexes evaluated by Protein–Ligand Interaction Profiler server. It is for the first time that used anticancer peptides of GP63 and the predicted and designed peptides.

Results

We used 3 peptides of GP63 based on the AntiCP 2.0 server with scores of 0.63, 0.53, and 0.49, and common peptides of GP63/MMP2 (continues peptide: mean the completely selected peptide after docking with non-anticancer effect, predicted with 0.58 score and designed peptides with 0.47 and 0.45 scores by AntiCP 2.0 server).

Conclusions

The antileishmanial and anticancer peptide research topics exemplify the multidisciplinary nature of peptide research. The advancement of therapeutics targeting cancer and/or Leishmania requires an interconnected research strategy shown in this work.

Background

The amount of stroma within the primary tumor is a prognostic parameter for colon cancer patients. This phenomenon can be assessed using the tumor–stroma ratio (TSR), which classifies tumors in stroma-low (≤50% stroma) and stroma-high (>50% stroma). Although the reproducibility for TSR determination is good, improvement might be expected from automation. The aim of this study was to investigate whether the scoring of the TSR in a semi- and fully automated method using deep learning algorithms is feasible.

Methods

A series of 75 colon cancer slides were selected from a trial series of the UNITED study. For the standard determination of the TSR, 3 observers scored the histological slides. Next, the slides were digitized, color normalized, and the stroma percentages were scored using semi- and fully automated deep learning algorithms. Correlations were determined using intraclass correlation coefficients (ICCs) and Spearman rank correlations.

Results

37 (49%) cases were classified as stroma-low and 38 (51%) as stroma-high by visual estimation. A high level of concordance between the 3 observers was reached, with ICCs of 0.91, 0.89, and 0.94 (all P < .001). Between visual and semi-automated assessment the ICC was 0.78 (95% CI 0.23–0.91, P-value 0.005), with a Spearman correlation of 0.88 (P < .001). Spearman correlation coefficients above 0.70 (N=3) were observed for visual estimation versus the fully automated scoring procedures.

Conclusion

Good correlations were observed between standard visual TSR determination and semi- and fully automated TSR scores. At this point, visual examination has the highest observer agreement, but semi-automated scoring could be helpful to support pathologists.

Diagnostic advances have not kept pace with the expansion of Lyme disease caused by Borrelia burgdorferi and transmitted by ticks. Lyme disease clinical manifestations can overlap with many other diagnoses making Lyme disease a critical part of many differential diagnoses in endemic areas. Current diagnostic blood tests rely on a 2-tiered algorithm for which the second step is either a time-consuming western blot or a whole cell lysate immunoassay. Neither of these second step tests allow for rapid results of this critical rule out test. We hypothesized that using western blot confirmation information, we could create computational models to propose recombinant second-tier tests that would allow for more rapid, automated, and specific testing algorithms. We propose here a framework for assessing retrospective data to determine putative recombinant assay components. A retrospective pediatric cohort of 2755 samples submitted for Lyme disease screening was assessed using support vector machine learning algorithms to optimize tier 1 diagnostic thresholds for the Vidas IgG II assay and determine optimal tier 2 components for both a positive and negative confirmation test. In cases where the tier 1 screen was negative, but clinical suspicion was high, we found that 1 protein (L58) could be used to reduce false-negative results. For second-tier testing of screen positive cases, we found that 6 proteins could be used to reduce false-positive results (L18, L39M, L39, L41, L45, and L58) with a final machine learning classifier or 2 proteins using a final rules-based approach (L41, L18). This led to an overall accuracy of 92.36% for the proposed algorithm without a final machine learning classifier and 92.12% with integration of the machine learning classifier in the final algorithm when compared to the IgG western blot as the gold-standard. Use of this framework across multiple assays and institutions will allow for a data-driven approach to assay development to provide laboratories and patients with the improvements in turnaround time needed for this testing.

Background and objective

Training a robust cancer diagnostic or prognostic artificial intelligent model using histology images requires a large number of representative cases with labels or annotations, which are difficult to obtain. The histology snapshots available in published papers or case reports can be used to enrich the training dataset. However, the magnifications of these invaluable snapshots are generally unknown, which limits their usage. Therefore, a robust magnification predictor is required for utilizing those diverse snapshot repositories consisting of different diseases. This paper presents a magnification prediction model named Hagnifinder for H&E-stained histological images.

Methods

Hagnifinder is a regression model based on a modified convolutional neural network (CNN) that contains 3 modules: Feature Extraction Module, Regression Module, and Adaptive Scaling Module (ASM). In the training phase, the Feature Extraction Module first extracts the image features. Secondly, the ASM is proposed to address the learned feature values uneven distribution problem. Finally, the Regression Module estimates the mapping between the regularized extracted features and the magnifications. We construct a new dataset for training a robust model, named Hagni40, consisting of 94 643 H&E-stained histology image patches at 40 different magnifications of 13 types of cancer based on The Cancer Genome Atlas. To verify the performance of the Hagnifinder, we measure the accuracy of the predictions by setting the maximum allowable difference values (0.5, 1, and 5) between the predicted magnification and the actual magnification. We compare Hagnifinder with state-of-the-art methods on a public dataset BreakHis and the Hagni40.

Results

The Hagnifinder provides consistent prediction accuracy, with a mean accuracy of 98.9%, across 40 different magnifications and 13 different cancer types when Resnet50 is used as the feature extractor. Compared with the state-of-the-art methods focusing on 4–5 levels of magnification classification, the Hagnifinder achieves the best and most comparable performance in the BreakHis and Hagni40 datasets.

Conclusions

The experimental results suggest that Hagnifinder can be a valuable tool for predicting the associated magnification of any given histology image.