NuCLS: A scalable crowdsourcing approach and dataset for nucleus classification and segmentation in breast cancer

Genomic characterization and histologic analysis of uterine leiomyosarcoma arising from leiomyoma with bizarre nuclei

Leiomyoma with bizarre nuclei (LM-BN) is a rare variant of leiomyoma with a benign clinical course. In contrast, leiomyosarcoma (LMS) is a high-grade, malignant neoplasm characterized by high recurrence rates and poor survival. While LM-BN and LMS show distinct morphologies, they share similar immunoprofiles and molecular alterations, with both considered 'genomically unstable'. Rare cases of LM-BN associated with LMS have been reported; however, the histogenesis and molecular relationship between these two tumors remains unclear. In this study, we assessed 11 cases of LMS arising in conjunction with LM-BN confirmed by histology and immunohistochemistry (IHC), further analyzed by clinical, histologic, and molecular characteristics of these distinct components. LM-BN and LMS had similar p16 and p53 IHC patterns, but LMS had a higher Ki-67 index and lower estrogen and progesterone eceptor expression. Digital image analysis based on cytologic features revealed spatial relationships between LMS and LM-BN. Genomic copy number alterations (CNAs) demonstrated the same clonal origin of LMS arising from existing LM-BN through conserved CNAs. LMS harbored highly complex CNAs and more frequent losses of the TP53, RB1, and PTEN genomic regions than LM-BN (p = 0.0031), with CDKN2A/B deletion identified in LMS only. Mutational profiling revealed many shared oncogenic alterations in both LM-BN and LMS; however, additional mutations were present within LMS, indicative of tumor progression through progressive genomic instability. Analysis of spatial transcriptomes defined uniquely expressed gene signatures that matched the geographic distribution of LM-BN, LMS, and other cell types. Our findings indicate for the first time that a subset of LMS arises from an existing LM-BN, and highly complex genomic alterations could be potential high risks associated with disease progression in LM-BN. © 2024 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Classifying tumour infiltrating lymphocytes in oral squamous cell carcinoma histopathology using joint learning framework

Oral squamous cell carcinoma (OSCC) is the most common form of oral cancer, with increasing global incidence and have poor prognosis. Tumour-infiltrating lymphocytes (TILs) are recognized as a key prognostic indicator and play a vital role in OSCC grading. However, current methods for TILs quantification are based on subjective visual assessments, leading to inter-observer variability and inconsistent diagnostic reproducibility. Only a few studies have been conducted in automating TILs quantification for OSCC, existing methods use score-based systems that focus only on tissue-level spatial analysis, overlooking essential cellular-level information and do not provide TILs infiltration subcategories required for determining OSCC grading. To address these limitations, we propose OralTILs-ViT, a novel joint representation learning framework that integrates cellular and tissue-level information. Our model employs two parallel encoders: one extracts cellular features from cellular density maps, while the other processes tissue features from H&E-stained tissue images. This dual-encoder approach enables OralTILs-ViT to capture complex tissue-cellular interactions, classifying TILs infiltration categories consistent with Broders' grading system-"Moderate to Marked", "Slight" and "None to Very Less." This approach reflects pathology practices and increases TILs classification accuracy. To generate cellular density maps, we introduce TILSeg-MobileViT, a multiclass segmentation model trained using a weakly supervised method, minimizing the need for manual annotation of cellular masks and overcoming the limitations of previous TILs assessment techniques. An extensive evaluation of our methodology demonstrates that OralTILs-ViT with the configuration (Adam, α = 0.001) outperforms existing approaches, achieving 96.37% accuracy, 96.34% precision, 96.37% recall, and a 96.35% F1 score. Furthermore, TOPSIS analysis confirms that our method ranks first across all TILs infiltration categories. In summary, our proposed methodology outperforms single modality-representation learning approaches for accurate and automated TILs classification.

The CrowdGleason dataset: Learning the Gleason grade from crowds and experts

BACKGROUND

Currently, prostate cancer (PCa) diagnosis relies on the human analysis of prostate biopsy Whole Slide Images (WSIs) using the Gleason score. Since this process is error-prone and time-consuming, recent advances in machine learning have promoted the use of automated systems to assist pathologists. Unfortunately, labeled datasets for training and validation are scarce due to the need for expert pathologists to provide ground-truth labels.

METHODS

This work introduces a new prostate histopathological dataset named CrowdGleason, which consists of 19,077 patches from 1045 WSIs with various Gleason grades. The dataset was annotated using a crowdsourcing protocol involving seven pathologists-in-training to distribute the labeling effort. To provide a baseline analysis, two crowdsourcing methods based on Gaussian Processes (GPs) were evaluated for Gleason grade prediction: SVGPCR, which learns a model from the CrowdGleason dataset, and SVGPMIX, which combines data from the public dataset SICAPv2 and the CrowdGleason dataset. The performance of these methods was compared with other crowdsourcing and expert label-based methods through comprehensive experiments.

RESULTS

The results demonstrate that our GP-based crowdsourcing approach outperforms other methods for aggregating crowdsourced labels (κ=0.7048±0.0207) for SVGPCR vs.(κ=0.6576±0.0086) for SVGP with majority voting). SVGPCR trained with crowdsourced labels performs better than GP trained with expert labels from SICAPv2 (κ=0.6583±0.0220) and outperforms most individual pathologists-in-training (mean κ=0.5432). Additionally, SVGPMIX trained with a combination of SICAPv2 and CrowdGleason achieves the highest performance on both datasets (κ=0.7814±0.0083 and κ=0.7276±0.0260).

CONCLUSION

The experiments show that the CrowdGleason dataset can be successfully used for training and validating supervised and crowdsourcing methods. Furthermore, the crowdsourcing methods trained on this dataset obtain competitive results against those using expert labels. Interestingly, the combination of expert and non-expert labels opens the door to a future of massive labeling by incorporating both expert and non-expert pathologist annotators.

Enhancing AI Research for Breast Cancer: A Comprehensive Review of Tumor-Infiltrating Lymphocyte Datasets

The field of immunology is fundamental to our understanding of the intricate dynamics of the tumor microenvironment. In particular, tumor-infiltrating lymphocyte (TIL) assessment emerges as essential aspect in breast cancer cases. To gain comprehensive insights, the quantification of TILs through computer-assisted pathology (CAP) tools has become a prominent approach, employing advanced artificial intelligence models based on deep learning techniques. The successful recognition of TILs requires the models to be trained, a process that demands access to annotated datasets. Unfortunately, this task is hampered not only by the scarcity of such datasets, but also by the time-consuming nature of the annotation phase required to create them. Our review endeavors to examine publicly accessible datasets pertaining to the TIL domain and thereby become a valuable resource for the TIL community. The overall aim of the present review is thus to make it easier to train and validate current and upcoming CAP tools for TIL assessment by inspecting and evaluating existing publicly available online datasets.

Cell Segmentation With Globally Optimized Boundaries (CSGO): A Deep Learning Pipeline for Whole-Cell Segmentation in Hematoxylin-and-Eosin-Stained Tissues

Accurate whole-cell segmentation is essential in various biomedical applications, particularly in studying the tumor microenvironment. Despite advancements in machine learning for nuclei segmentation in hematoxylin and eosin (H&E)-stained images, there remains a need for effective whole-cell segmentation methods. This study aimed to develop a deep learning-based pipeline to automatically segment cells in H&E-stained tissues, thereby advancing the capabilities of pathological image analysis. The Cell Segmentation with Globally Optimized boundaries (CSGO) framework integrates nuclei and membrane segmentation algorithms, followed by postprocessing using an energy-based watershed method. Specifically, we used the You Only Look Once (YOLO) object detection algorithm for nuclei segmentation and U-Net for membrane segmentation. The membrane detection model was trained on a data set of 7 hepatocellular carcinomas and 11 normal liver tissue patches. The cell segmentation performance was extensively evaluated on 5 external data sets, including liver, lung, and oral disease cases. CSGO demonstrated superior performance over the state-of-the-art method Cellpose, achieving higher F1 scores ranging from 0.37 to 0.53 at an intersection over union threshold of 0.5 in 4 of the 5 external datasets, compared to that of Cellpose from 0.21 to 0.36. These results underscore the robustness and accuracy of our approach in various tissue types. A web-based application is available at https://ai.swmed.edu/projects/csgo, providing a user-friendly platform for researchers to apply our method to their own data sets. Our method exhibits remarkable versatility in whole-cell segmentation across diverse cancer subtypes, serving as an accurate and reliable tool to facilitate tumor microenvironment studies. The advancements presented in this study have the potential to significantly enhance the precision and efficiency of pathologic image analysis, contributing to better understanding and treatment of cancer.

WSSS-CRAM: precise segmentation of histopathological images via class region activation mapping

Introduction

Fast, accurate, and automatic analysis of histopathological images using digital image processing and deep learning technology is a necessary task. Conventional histopathological image analysis algorithms require the manual design of features, while deep learning methods can achieve fast prediction and accurate analysis, but rely on the drive of a large amount of labeled data.

Methods

In this work, we introduce WSSS-CRAM a weakly-supervised semantic segmentation that can obtain detailed pixel-level labels from image-level annotated data. Specifically, we use a discriminative activation strategy to generate category-specific image activation maps via class labels. The category-specific activation maps are then post-processed using conditional random fields to obtain reliable regions that are directly used as ground-truth labels for the segmentation branch. Critically, the two steps of the pseudo-label acquisition and training segmentation model are integrated into an end-to-end model for joint training in this method.

Results

Through quantitative evaluation and visualization results, we demonstrate that the framework can predict pixel-level labels from image-level labels, and also perform well when testing images without image-level annotations.

Discussion

Future, we consider extending the algorithm to different pathological datasets and types of tissue images to validate its generalization capability.

Quantitative Modeling to Characterize Maternal Inflammatory Response of Histologic Chorioamnionitis in Placental Membranes

PROBLEM

The placental membranes are a key barrier to fetal and uterine infection. Inflammation of the membranes, diagnosed as maternal inflammatory response (MIR) or alternatively as acute chorioamnionitis, is associated with adverse maternal-fetal outcomes. MIR is staged 1-3, with higher stages indicating more hazardous inflammation. However, the diagnosis relies upon subjective evaluation and has not been deeply characterized. The goal of this work is to develop a cell classifier for eight placental membrane cells and quantitatively characterize MIR1-2.

METHOD OF STUDY

Hematoxylin and eosin (H&E)-stained placental membrane slides were digitized. A convolutional neural network was trained on a dataset of hand-annotated and machine learning-identified cells. Overall cell class-level metrics were calculated. The model was applied to 20 control, 20 MIR1, and 23 MIR2 placental membrane cases. MIR cell composition and neutrophil distribution were assessed via density and Ripley's cross K-function. Clinical data were compared to neutrophil density and distribution.

RESULTS

The classification model achieved a test-set accuracy of 0.845, with high precision and recall for amniocytes, decidual cells, endothelial cells, and trophoblasts. Using this model to classify 53 073 cells from healthy and MIR1-2 placental membranes, we found that (1) MIR1-2 have higher neutrophil density and fewer decidual cells and trophoblasts, (2) Neutrophils colocalize heavily around decidual cells in healthy placental membranes and around trophoblasts in MIR1, (3) Neutrophil density impacts distribution in MIR, and (4) Neutrophil metrics correlate with features of clinical chorioamnionitis.

CONCLUSIONS

This paper introduces cell classification into the placental membranes and quantifies cell composition and neutrophil spatial distributions in MIR.

A Quantitative Measurement Method for Nuclear-Pleomorphism Scoring in Breast Cancer

BACKGROUND/OBJECTIVES

Nuclear pleomorphism, a crucial determinant of breast cancer grading under the Nottingham Histopathology Grading (NHG) system, remains inadequately quantified in the existing literature. Motivated by this gap, our study seeks to investigate and establish correlations among morphological features across various scores of nuclear pleomorphism, as per the NHG system. We aim to quantify nuclear pleomorphism across these scores and validate our proposed measurement method against ground-truth data.

METHODS

Initially, we deconstruct the descriptions of nuclear pleomorphism into three core elements: size, shape, and appearance. These elements are subsequently mathematically modeled into equations, termed ESize, EShape, and EAppearance. These equations are then integrated into a unified model termed Harmonic Mean (HM). The HM equation yields a value approaching 1 for nuclei demonstrating characteristics of score-3 nuclear pleomorphism and near 0 for those exhibiting features of score-1 nuclear pleomorphism.

RESULTS

The proposed HM model demonstrates promising performance metrics, including Accuracy, Recall, Specificity, Precision, and F1-score, with values of 0.97, 0.96, 0.97, 0.94, and 0.95, respectively.

CONCLUSIONS

In summary, this study proposes the HM equation as a novel feature for the precise quantification of nuclear pleomorphism in breast cancer.

Computational methods and biomarker discovery strategies for spatial proteomics: a review in immuno-oncology

Advancements in imaging technologies have revolutionized our ability to deeply profile pathological tissue architectures, generating large volumes of imaging data with unparalleled spatial resolution. This type of data collection, namely, spatial proteomics, offers invaluable insights into various human diseases. Simultaneously, computational algorithms have evolved to manage the increasing dimensionality of spatial proteomics inherent in this progress. Numerous imaging-based computational frameworks, such as computational pathology, have been proposed for research and clinical applications. However, the development of these fields demands diverse domain expertise, creating barriers to their integration and further application. This review seeks to bridge this divide by presenting a comprehensive guideline. We consolidate prevailing computational methods and outline a roadmap from image processing to data-driven, statistics-informed biomarker discovery. Additionally, we explore future perspectives as the field moves toward interfacing with other quantitative domains, holding significant promise for precision care in immuno-oncology.

A panoptic segmentation dataset and deep-learning approach for explainable scoring of tumor-infiltrating lymphocytes

Tumor-Infiltrating Lymphocytes (TILs) have strong prognostic and predictive value in breast cancer, but their visual assessment is subjective. To improve reproducibility, the International Immuno-oncology Working Group recently released recommendations for the computational assessment of TILs that build on visual scoring guidelines. However, existing resources do not adequately address these recommendations due to the lack of annotation datasets that enable joint, panoptic segmentation of tissue regions and cells. Moreover, existing deep-learning methods focus entirely on either tissue segmentation or cell nuclei detection, which complicates the process of TILs assessment by necessitating the use of multiple models and reconciling inconsistent predictions. We introduce PanopTILs, a region and cell-level annotation dataset containing 814,886 nuclei from 151 patients, openly accessible at: sites.google.com/view/panoptils . Using PanopTILs we developed MuTILs, a neural network optimized for assessing TILs in accordance with clinical recommendations. MuTILs is a concept bottleneck model designed to be interpretable and to encourage sensible predictions at multiple resolutions. Using a rigorous internal-external cross-validation procedure, MuTILs achieves an AUROC of 0.93 for lymphocyte detection and a DICE coefficient of 0.81 for tumor-associated stroma segmentation. Our computational score closely matched visual scores from 2 pathologists (Spearman R = 0.58-0.61, p < 0.001). Moreover, computational TILs scores had a higher prognostic value than visual scores, independent of TNM stage and patient age. In conclusion, we introduce a comprehensive open data resource and a modeling approach for detailed mapping of the breast tumor microenvironment.

Inconsistency between Human Observation and Deep Learning Models: Assessing Validity of Postmortem Computed Tomography Diagnosis of Drowning

Drowning diagnosis is a complicated process in the autopsy, even with the assistance of autopsy imaging and the on-site information from where the body was found. Previous studies have developed well-performed deep learning (DL) models for drowning diagnosis. However, the validity of the DL models was not assessed, raising doubts about whether the learned features accurately represented the medical findings observed by human experts. In this paper, we assessed the medical validity of DL models that had achieved high classification performance for drowning diagnosis. This retrospective study included autopsy cases aged 8-91 years who underwent postmortem computed tomography between 2012 and 2021 (153 drowning and 160 non-drowning cases). We first trained three deep learning models from a previous work and generated saliency maps that highlight important features in the input. To assess the validity of models, pixel-level annotations were created by four radiological technologists and further quantitatively compared with the saliency maps. All the three models demonstrated high classification performance with areas under the receiver operating characteristic curves of 0.94, 0.97, and 0.98, respectively. On the other hand, the assessment results revealed unexpected inconsistency between annotations and models' saliency maps. In fact, each model had, respectively, around 30%, 40%, and 80% of irrelevant areas in the saliency maps, suggesting the predictions of the DL models might be unreliable. The result alerts us in the careful assessment of DL tools, even those with high classification performance.

VOLTA: an enVironment-aware cOntrastive ceLl represenTation leArning for histopathology

In clinical oncology, many diagnostic tasks rely on the identification of cells in histopathology images. While supervised machine learning techniques necessitate the need for labels, providing manual cell annotations is time-consuming. In this paper, we propose a self-supervised framework (enVironment-aware cOntrastive cell represenTation learning: VOLTA) for cell representation learning in histopathology images using a technique that accounts for the cell's mutual relationship with its environment. We subject our model to extensive experiments on data collected from multiple institutions comprising over 800,000 cells and six cancer types. To showcase the potential of our proposed framework, we apply VOLTA to ovarian and endometrial cancers and demonstrate that our cell representations can be utilized to identify the known histotypes of ovarian cancer and provide insights that link histopathology and molecular subtypes of endometrial cancer. Unlike supervised models, we provide a framework that can empower discoveries without any annotation data, even in situations where sample sizes are limited.

Reproducible Reporting of the Collection and Evaluation of Annotations for Artificial Intelligence Models

This work puts forth and demonstrates the utility of a reporting framework for collecting and evaluating annotations of medical images used for training and testing artificial intelligence (AI) models in assisting detection and diagnosis. AI has unique reporting requirements, as shown by the AI extensions to the Consolidated Standards of Reporting Trials (CONSORT) and Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklists and the proposed AI extensions to the Standards for Reporting Diagnostic Accuracy (STARD) and Transparent Reporting of a Multivariable Prediction model for Individual Prognosis or Diagnosis (TRIPOD) checklists. AI for detection and/or diagnostic image analysis requires complete, reproducible, and transparent reporting of the annotations and metadata used in training and testing data sets. In an earlier work by other researchers, an annotation workflow and quality checklist for computational pathology annotations were proposed. In this manuscript, we operationalize this workflow into an evaluable quality checklist that applies to any reader-interpreted medical images, and we demonstrate its use for an annotation effort in digital pathology. We refer to this quality framework as the Collection and Evaluation of Annotations for Reproducible Reporting of Artificial Intelligence (CLEARR-AI).

Learning from crowds for automated histopathological image segmentation

Automated semantic segmentation of histopathological images is an essential task in Computational Pathology (CPATH). The main limitation of Deep Learning (DL) to address this task is the scarcity of expert annotations. Crowdsourcing (CR) has emerged as a promising solution to reduce the individual (expert) annotation cost by distributing the labeling effort among a group of (non-expert) annotators. Extracting knowledge in this scenario is challenging, as it involves noisy annotations. Jointly learning the underlying (expert) segmentation and the annotators' expertise is currently a commonly used approach. Unfortunately, this approach is frequently carried out by learning a different neural network for each annotator, which scales poorly when the number of annotators grows. For this reason, this strategy cannot be easily applied to real-world CPATH segmentation. This paper proposes a new family of methods for CR segmentation of histopathological images. Our approach consists of two coupled networks: a segmentation network (for learning the expert segmentation) and an annotator network (for learning the annotators' expertise). We propose to estimate the annotators' behavior with only one network that receives the annotator ID as input, achieving scalability on the number of annotators. Our family is composed of three different models for the annotator network. Within this family, we propose a novel modeling of the annotator network in the CR segmentation literature, which considers the global features of the image. We validate our methods on a real-world dataset of Triple Negative Breast Cancer images labeled by several medical students. Our new CR modeling achieves a Dice coefficient of 0.7827, outperforming the well-known STAPLE (0.7039) and being competitive with the supervised method with expert labels (0.7723). The code is available at https://github.com/wizmik12/CRowd_Seg.

Microsnoop: A generalist tool for microscopy image representation

Accurate profiling of microscopy images from small scale to high throughput is an essential procedure in basic and applied biological research. Here, we present Microsnoop, a novel deep learning-based representation tool trained on large-scale microscopy images using masked self-supervised learning. Microsnoop can process various complex and heterogeneous images, and we classified images into three categories: single-cell, full-field, and batch-experiment images. Our benchmark study on 10 high-quality evaluation datasets, containing over 2,230,000 images, demonstrated Microsnoop's robust and state-of-the-art microscopy image representation ability, surpassing existing generalist and even several custom algorithms. Microsnoop can be integrated with other pipelines to perform tasks such as superresolution histopathology image and multimodal analysis. Furthermore, Microsnoop can be adapted to various hardware and can be easily deployed on local or cloud computing platforms. We will regularly retrain and reevaluate the model using community-contributed data to consistently improve Microsnoop.

A population-level digital histologic biomarker for enhanced prognosis of invasive breast cancer

Breast cancer is a heterogeneous disease with variable survival outcomes. Pathologists grade the microscopic appearance of breast tissue using the Nottingham criteria, which are qualitative and do not account for noncancerous elements within the tumor microenvironment. Here we present the Histomic Prognostic Signature (HiPS), a comprehensive, interpretable scoring of the survival risk incurred by breast tumor microenvironment morphology. HiPS uses deep learning to accurately map cellular and tissue structures to measure epithelial, stromal, immune, and spatial interaction features. It was developed using a population-level cohort from the Cancer Prevention Study-II and validated using data from three independent cohorts, including the Prostate, Lung, Colorectal, and Ovarian Cancer trial, Cancer Prevention Study-3, and The Cancer Genome Atlas. HiPS consistently outperformed pathologists in predicting survival outcomes, independent of tumor-node-metastasis stage and pertinent variables. This was largely driven by stromal and immune features. In conclusion, HiPS is a robustly validated biomarker to support pathologists and improve patient prognosis.

Annotation protocol and crowdsourcing multiple instance learning classification of skin histological images: The CR-AI4SkIN dataset

Digital Pathology (DP) has experienced a significant growth in recent years and has become an essential tool for diagnosing and prognosis of tumors. The availability of Whole Slide Images (WSIs) and the implementation of Deep Learning (DL) algorithms have paved the way for the appearance of Artificial Intelligence (AI) systems that support the diagnosis process. These systems require extensive and varied data for their training to be successful. However, creating labeled datasets in histopathology is laborious and time-consuming. We have developed a crowdsourcing-multiple instance labeling/learning protocol that is applied to the creation and use of the CR-AI4SkIN dataset.2 CR-AI4SkIN contains 271 WSIs of 7 Cutaneous Spindle Cell (CSC) neoplasms with expert and non-expert labels at region and WSI levels. It is the first dataset of these types of neoplasms made available. The regions selected by the experts are used to learn an automatic extractor of Regions of Interest (ROIs) from WSIs. To produce the embedding of each WSI, the representations of patches within the ROIs are obtained using a contrastive learning method, and then combined. Finally, they are fed to a Gaussian process-based crowdsourcing classifier, which utilizes the noisy non-expert WSI labels. We validate our crowdsourcing-multiple instance learning method in the CR-AI4SkIN dataset, addressing a binary classification problem (malign vs. benign). The proposed method obtains an F1 score of 0.7911 on the test set, outperforming three widely used aggregation methods for crowdsourcing tasks. Furthermore, our crowdsourcing method also outperforms the supervised model with expert labels on the test set (F1-score = 0.6035). The promising results support the proposed crowdsourcing multiple instance learning annotation protocol. It also validates the automatic extraction of interest regions and the use of contrastive embedding and Gaussian process classification to perform crowdsourcing classification tasks.

Democratizing Pathological Image Segmentation with Lay Annotators via Molecular-empowered Learning

Multi-class cell segmentation in high-resolution Giga-pixel whole slide images (WSI) is critical for various clinical applications. Training such an AI model typically requires labor-intensive pixel-wise manual annotation from experienced domain experts (e.g., pathologists). Moreover, such annotation is error-prone when differentiating fine-grained cell types (e.g., podocyte and mesangial cells) via the naked human eye. In this study, we assess the feasibility of democratizing pathological AI deployment by only using lay annotators (annotators without medical domain knowledge). The contribution of this paper is threefold: (1) We proposed a molecular-empowered learning scheme for multi-class cell segmentation using partial labels from lay annotators; (2) The proposed method integrated Giga-pixel level molecular-morphology cross-modality registration, molecular-informed annotation, and molecular-oriented segmentation model, so as to achieve significantly superior performance via 3 lay annotators as compared with 2 experienced pathologists; (3) A deep corrective learning (learning with imperfect label) method is proposed to further improve the segmentation performance using partially annotated noisy data. From the experimental results, our learning method achieved F1 = 0.8496 using molecular-informed annotations from lay annotators, which is better than conventional morphology-based annotations (F1 = 0.7015) from experienced pathologists. Our method democratizes the development of a pathological segmentation deep model to the lay annotator level, which consequently scales up the learning process similar to a non-medical computer vision task. The official implementation and cell annotations are publicly available at https://github.com/hrlblab/MolecularEL.

Pitfalls in machine learning-based assessment of tumor-infiltrating lymphocytes in breast cancer: A report of the International Immuno-Oncology Biomarker Working Group on Breast Cancer

The clinical significance of the tumor-immune interaction in breast cancer is now established, and tumor-infiltrating lymphocytes (TILs) have emerged as predictive and prognostic biomarkers for patients with triple-negative (estrogen receptor, progesterone receptor, and HER2-negative) breast cancer and HER2-positive breast cancer. How computational assessments of TILs might complement manual TIL assessment in trial and daily practices is currently debated. Recent efforts to use machine learning (ML) to automatically evaluate TILs have shown promising results. We review state-of-the-art approaches and identify pitfalls and challenges of automated TIL evaluation by studying the root cause of ML discordances in comparison to manual TIL quantification. We categorize our findings into four main topics: (1) technical slide issues, (2) ML and image analysis aspects, (3) data challenges, and (4) validation issues. The main reason for discordant assessments is the inclusion of false-positive areas or cells identified by performance on certain tissue patterns or design choices in the computational implementation. To aid the adoption of ML for TIL assessment, we provide an in-depth discussion of ML and image analysis, including validation issues that need to be considered before reliable computational reporting of TILs can be incorporated into the trial and routine clinical management of patients with triple-negative breast cancer. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Deep-Fuzz: A synergistic integration of deep learning and fuzzy water flows for fine-grained nuclei segmentation in digital pathology

Robust semantic segmentation of tumour micro-environment is one of the major open challenges in machine learning enabled computational pathology. Though deep learning based systems have made significant progress, their task agnostic data driven approach often lacks the contextual grounding necessary in biomedical applications. We present a novel fuzzy water flow scheme that takes the coarse segmentation output of a base deep learning framework to then provide a more fine-grained and instance level robust segmentation output. Our two stage synergistic segmentation method, Deep-Fuzz, works especially well for overlapping objects, and achieves state-of-the-art performance in four public cell nuclei segmentation datasets. We also show through visual examples how our final output is better aligned with pathological insights, and thus more clinically interpretable.

Panoptic quality should be avoided as a metric for assessing cell nuclei segmentation and classification in digital pathology

Panoptic Quality (PQ), designed for the task of "Panoptic Segmentation" (PS), has been used in several digital pathology challenges and publications on cell nucleus instance segmentation and classification (ISC) since its introduction in 2019. Its purpose is to encompass the detection and the segmentation aspects of the task in a single measure, so that algorithms can be ranked according to their overall performance. A careful analysis of the properties of the metric, its application to ISC and the characteristics of nucleus ISC datasets, shows that is not suitable for this purpose and should be avoided. Through a theoretical analysis we demonstrate that PS and ISC, despite their similarities, have some fundamental differences that make PQ unsuitable. We also show that the use of the Intersection over Union as a matching rule and as a segmentation quality measure within PQ is not adapted for such small objects as nuclei. We illustrate these findings with examples taken from the NuCLS and MoNuSAC datasets. The code for replicating our results is available on GitHub ( https://github.com/adfoucart/panoptic-quality-suppl ).

Machine learning classification of placental villous infarction, perivillous fibrin deposition, and intervillous thrombus

INTRODUCTION

Placental parenchymal lesions are commonly encountered and carry significant clinical associations. However, they are frequently missed or misclassified by general practice pathologists. Interpretation of pathology slides has emerged as one of the most successful applications of machine learning (ML) in medicine with applications ranging from cancer detection and prognostication to transplant medicine. The goal of this study was to use a whole-slide learning model to identify and classify placental parenchymal lesions including villous infarctions, intervillous thrombi (IVT), and perivillous fibrin deposition (PVFD).

METHODS

We generated whole slide images from placental discs examined at our institution with infarct, IVT, PVFD, or no macroscopic lesion. Slides were analyzed as a set of overlapping patches. We extracted feature vectors from each patch using a pretrained convolutional neural network (EfficientNetV2L). We trained a model to assign attention to each vector and used the attentions as weights to produce a pooled feature vector. The pooled vector was classified as normal or 1 of 3 lesions using a fully connected network. Patch attention was plotted to highlight informative areas of the slide.

RESULTS

Overall balanced accuracy in a test set of held-out slides was 0.86 with receiver-operator characteristic areas under the curve of 0.917-0.993. Cases of PVFD were frequently miscalled as normal or infarcts, the latter possibly due to the perivillous fibrin found at the periphery of infarctions. We used attention maps to further understand some errors, including one most likely due to poor tissue fixation and processing.

DISCUSSION

We used a whole-slide learning paradigm to train models to recognize three of the most common placental parenchymal lesions. We used attention maps to gain insight into model function, which differed from intuitive explanations.

A pathomic approach for tumor-infiltrating lymphocytes classification on breast cancer digital pathology images

Background and objectives

The detection of tumor-infiltrating lymphocytes (TILs) could aid in the development of objective measures of the infiltration grade and can support decision-making in breast cancer (BC). However, manual quantification of TILs in BC histopathological whole slide images (WSI) is currently based on a visual assessment, thus resulting not standardized, not reproducible, and time-consuming for pathologists. In this work, a novel pathomic approach, aimed to apply high-throughput image feature extraction techniques to analyze the microscopic patterns in WSI, is proposed. In fact, pathomic features provide additional information concerning the underlying biological processes compared to the WSI visual interpretation, thus providing more easily interpretable and explainable results than the most frequently investigated Deep Learning based methods in the literature.

Methods

A dataset containing 1037 regions of interest with tissue compartments and TILs annotated on 195 TNBC and HER2+ BC hematoxylin and eosin (H&E)-stained WSI was used. After segmenting nuclei within tumor-associated stroma using a watershed-based approach, 71 pathomic features were extracted from each nucleus and reduced using a Spearman's correlation filter followed by a nonparametric Wilcoxon rank-sum test and least absolute shrinkage and selection operator. The relevant features were used to classify each candidate nucleus as either TILs or non-TILs using 5 multivariable machine learning classification models trained using 5-fold cross-validation (1) without resampling, (2) with the synthetic minority over-sampling technique and (3) with downsampling. The prediction performance of the models was assessed using ROC curves.

Results

21 features were selected, with most of them related to the well-known TILs properties of having regular shape, clearer margins, high peak intensity, more homogeneous enhancement and different textural pattern than other cells. The best performance was obtained by Random-Forest with ROC AUC of 0.86, regardless of resampling technique.

Conclusions

The presented approach holds promise for the classification of TILs in BC H&E-stained WSI and could provide support to pathologists for a reliable, rapid and interpretable clinical assessment of TILs in BC.

Multi-modality artificial intelligence in digital pathology

In common medical procedures, the time-consuming and expensive nature of obtaining test results plagues doctors and patients. Digital pathology research allows using computational technologies to manage data, presenting an opportunity to improve the efficiency of diagnosis and treatment. Artificial intelligence (AI) has a great advantage in the data analytics phase. Extensive research has shown that AI algorithms can produce more up-to-date and standardized conclusions for whole slide images. In conjunction with the development of high-throughput sequencing technologies, algorithms can integrate and analyze data from multiple modalities to explore the correspondence between morphological features and gene expression. This review investigates using the most popular image data, hematoxylin-eosin stained tissue slide images, to find a strategic solution for the imbalance of healthcare resources. The article focuses on the role that the development of deep learning technology has in assisting doctors' work and discusses the opportunities and challenges of AI.

NuCLS: A scalable crowdsourcing approach and dataset for nucleus classification and segmentation in breast cancer

BACKGROUND

Deep learning enables accurate high-resolution mapping of cells and tissue structures that can serve as the foundation of interpretable machine-learning models for computational pathology. However, generating adequate labels for these structures is a critical barrier, given the time and effort required from pathologists.

RESULTS

This article describes a novel collaborative framework for engaging crowds of medical students and pathologists to produce quality labels for cell nuclei. We used this approach to produce the NuCLS dataset, containing >220,000 annotations of cell nuclei in breast cancers. This builds on prior work labeling tissue regions to produce an integrated tissue region- and cell-level annotation dataset for training that is the largest such resource for multi-scale analysis of breast cancer histology. This article presents data and analysis results for single and multi-rater annotations from both non-experts and pathologists. We present a novel workflow that uses algorithmic suggestions to collect accurate segmentation data without the need for laborious manual tracing of nuclei. Our results indicate that even noisy algorithmic suggestions do not adversely affect pathologist accuracy and can help non-experts improve annotation quality. We also present a new approach for inferring truth from multiple raters and show that non-experts can produce accurate annotations for visually distinctive classes.

CONCLUSIONS

This study is the most extensive systematic exploration of the large-scale use of wisdom-of-the-crowd approaches to generate data for computational pathology applications.