The future of artificial intelligence in digital pathology - results of a survey across stakeholder groups

Assessment of PD-L1 expression and tumour infiltrating lymphocytes in early-stage non-small cell lung carcinoma with artificial intelligence algorithms

AIMS

To study programmed death ligand 1 (PD-L1) expression and tumour infiltrating lymphocytes (TILs) in patients with early-stage non-small cell lung carcinoma (NSCLC) with artificial intelligence (AI) algorithms.

METHODS

The study included samples from 50 early-stage NSCLCs. PD-L1 immunohistochemistry (IHC) stained slides (clone SP263) were scored manually and with two different AI tools (PathAI and Navify Digital Pathology) by three pathologists. TILs were digitally assessed on H&E and CD8 IHC stained sections with two different algorithms (PathAI and Navify Digital Pathology, respectively). The agreement between observers and methods for each biomarker was analysed. For PD-L1, the turn-around time (TAT) for manual versus AI-assisted scoring was recorded.

RESULTS

Agreement was higher in tumours with low PD-L1 expression regardless of the approach. Both AI-powered tools identified a significantly higher number of cases equal or above 1% PD-L1 tumour proportion score as compared with manual scoring (p=0.00015), a finding with potential therapeutic implications. Regarding TAT, there were significant differences between manual scoring and AI use (p value <0.0001 for all comparisons). The total TILs density with the PathAI algorithm and the total density of CD8+ cells with the Navify Digital Pathology software were significantly correlated (τ=0.49 (95% CI 0.37, 0.61), p value<0.0001).

CONCLUSIONS

This preliminary study supports the use of AI algorithms for the scoring of PD-L1 and TILs in patients with NSCLC.

Automated curation of large-scale cancer histopathology image datasets using deep learning

BACKGROUND

Artificial intelligence (AI) has numerous applications in pathology, supporting diagnosis and prognostication in cancer. However, most AI models are trained on highly selected data, typically one tissue slide per patient. In reality, especially for large surgical resection specimens, dozens of slides can be available for each patient. Manually sorting and labelling whole-slide images (WSIs) is a very time-consuming process, hindering the direct application of AI on the collected tissue samples from large cohorts. In this study we addressed this issue by developing a deep-learning (DL)-based method for automatic curation of large pathology datasets with several slides per patient.

METHODS

We collected multiple large multicentric datasets of colorectal cancer histopathological slides from the United Kingdom (FOXTROT, N = 21,384 slides; CR07, N = 7985 slides) and Germany (DACHS, N = 3606 slides). These datasets contained multiple types of tissue slides, including bowel resection specimens, endoscopic biopsies, lymph node resections, immunohistochemistry-stained slides, and tissue microarrays. We developed, trained, and tested a deep convolutional neural network model to predict the type of slide from the slide overview (thumbnail) image. The primary statistical endpoint was the macro-averaged area under the receiver operating curve (AUROCs) for detection of the type of slide.

RESULTS

In the primary dataset (FOXTROT), with an AUROC of 0.995 [95% confidence interval [CI]: 0.994-0.996] the algorithm achieved a high classification performance and was able to accurately predict the type of slide from the thumbnail image alone. In the two external test cohorts (CR07, DACHS) AUROCs of 0.982 [95% CI: 0.979-0.985] and 0.875 [95% CI: 0.864-0.887] were observed, which indicates the generalizability of the trained model on unseen datasets. With a confidence threshold of 0.95, the model reached an accuracy of 94.6% (7331 classified cases) in CR07 and 85.1% (2752 classified cases) for the DACHS cohort.

CONCLUSION

Our findings show that using the low-resolution thumbnail image is sufficient to accurately classify the type of slide in digital pathology. This can support researchers to make the vast resource of existing pathology archives accessible to modern AI models with only minimal manual annotations.

The current state of digital cytology and artificial intelligence (AI): global survey results from the American Society of Cytopathology Digital Cytology Task Force

INTRODUCTION

The integration of whole slide imaging (WSI) and artificial intelligence (AI) with digital cytology has been growing gradually. Therefore, there is a need to evaluate the current state of digital cytology. This study aimed to determine the current landscape of digital cytology via a survey conducted as part of the American Society of Cytopathology (ASC) Digital Cytology White Paper Task Force.

MATERIALS AND METHODS

A survey with 43 questions pertaining to the current practices and experiences of WSI and AI in both surgical pathology and cytology was created. The survey was sent to members of the ASC, the International Academy of Cytology (IAC), and the Papanicolaou Society of Cytopathology (PSC). Responses were recorded and analyzed.

RESULTS

In total, 327 individuals participated in the survey, spanning a diverse array of practice settings, roles, and experiences around the globe. The majority of responses indicated there was routine scanning of surgical pathology slides (n = 134; 61%) with fewer respondents scanning cytology slides (n = 150; 46%). The primary challenge for surgical WSI is the need for faster scanning and cost minimization, whereas image quality is the top issue for cytology WSI. AI tools are not widely utilized, with only 16% of participants using AI for surgical pathology samples and 13% for cytology practice.

CONCLUSIONS

Utilization of digital pathology is limited in cytology laboratories as compared to surgical pathology. However, as more laboratories are willing to implement digital cytology in the near future, the establishment of practical clinical guidelines is needed.

Cutting-edge technology and automation in the pathology laboratory

One of the goals of pathology is to standardize laboratory practices to increase the precision and effectiveness of diagnostic testing, which will ultimately enhance patient care and results. Standardization is crucial in the domains of tissue processing, analysis, and reporting. To enhance diagnostic testing, innovative technologies are also being created and put into use. Furthermore, although problems like algorithm training and data privacy issues still need to be resolved, digital pathology and artificial intelligence are emerging in a structured manner. Overall, for the field of pathology to advance and for patient care to be improved, standard laboratory practices and innovative technologies must be adopted. In this paper, we describe the state-of-the-art of automation in pathology laboratories in order to lead technological progress and evolution. By anticipating laboratory needs and demands, the aim is to inspire innovation tools and processes as positively transformative support for operators, organizations, and patients.

Towards a general-purpose foundation model for computational pathology

Quantitative evaluation of tissue images is crucial for computational pathology (CPath) tasks, requiring the objective characterization of histopathological entities from whole-slide images (WSIs). The high resolution of WSIs and the variability of morphological features present significant challenges, complicating the large-scale annotation of data for high-performance applications. To address this challenge, current efforts have proposed the use of pretrained image encoders through transfer learning from natural image datasets or self-supervised learning on publicly available histopathology datasets, but have not been extensively developed and evaluated across diverse tissue types at scale. We introduce UNI, a general-purpose self-supervised model for pathology, pretrained using more than 100 million images from over 100,000 diagnostic H&E-stained WSIs (>77 TB of data) across 20 major tissue types. The model was evaluated on 34 representative CPath tasks of varying diagnostic difficulty. In addition to outperforming previous state-of-the-art models, we demonstrate new modeling capabilities in CPath such as resolution-agnostic tissue classification, slide classification using few-shot class prototypes, and disease subtyping generalization in classifying up to 108 cancer types in the OncoTree classification system. UNI advances unsupervised representation learning at scale in CPath in terms of both pretraining data and downstream evaluation, enabling data-efficient artificial intelligence models that can generalize and transfer to a wide range of diagnostically challenging tasks and clinical workflows in anatomic pathology.

Standardized Classification of Lung Adenocarcinoma Subtypes and Improvement of Grading Assessment Through Deep Learning

The histopathologic distinction of lung adenocarcinoma (LADC) subtypes is subject to high interobserver variability, which can compromise the optimal assessment of patient prognosis. Therefore, this study developed convolutional neural networks capable of distinguishing LADC subtypes and predicting disease-specific survival, according to the recently established LADC tumor grades. Consensus LADC histopathologic images were obtained from 17 expert pulmonary pathologists and one pathologist in training. Two deep learning models (AI-1 and AI-2) were trained to predict eight different LADC classes. Furthermore, the trained models were tested on an independent cohort of 133 patients. The models achieved high precision, recall, and F1 scores exceeding 0.90 for most of the LADC classes. Clear stratification of the three LADC grades was reached in predicting the disease-specific survival by the two models, with both Kaplan-Meier curves showing significance (P = 0.0017 and 0.0003). Moreover, both trained models showed high stability in the segmentation of each pair of predicted grades with low variation in the hazard ratio across 200 bootstrapped samples. These findings indicate that the trained convolutional neural networks improve the diagnostic accuracy of the pathologist and refine LADC grade assessment. Thus, the trained models are promising tools that may assist in the routine evaluation of LADC subtypes and grades in clinical practice.

Prediction models for hormone receptor status in female breast cancer do not extend to males: further evidence of sex-based disparity in breast cancer

Breast cancer prognosis and management for both men and women are reliant upon estrogen receptor alpha (ERα) and progesterone receptor (PR) expression to inform therapy. Previous studies have shown that there are sex-specific binding characteristics of ERα and PR in breast cancer and, counterintuitively, ERα expression is more common in male than female breast cancer. We hypothesized that these differences could have morphological manifestations that are undetectable to human observers but could be elucidated computationally. To investigate this, we trained attention-based multiple instance learning prediction models for ERα and PR using H&E-stained images of female breast cancer from the Cancer Genome Atlas (TCGA) (n = 1085) and deployed them on external female (n = 192) and male breast cancer images (n = 245). Both targets were predicted in the internal (AUROC for ERα prediction: 0.86 ± 0.02, p < 0.001; AUROC for PR prediction = 0.76 ± 0.03, p < 0.001) and external female cohorts (AUROC for ERα prediction: 0.78 ± 0.03, p < 0.001; AUROC for PR prediction = 0.80 ± 0.04, p < 0.001) but not the male cohort (AUROC for ERα prediction: 0.66 ± 0.14, p = 0.43; AUROC for PR prediction = 0.63 ± 0.04, p = 0.05). This suggests that subtle morphological differences invisible upon visual inspection may exist between the sexes, supporting previous immunohistochemical, genomic, and transcriptomic analyses.

Results of the European Society of Toxicologic Pathology Survey on the Use of Artificial Intelligence in Toxicologic Pathology

The European Society of Toxicologic Pathology (ESTP) initiated a survey through its Pathology 2.0 workstream in partnership with sister professional societies in Europe and North America to generate a snapshot of artificial intelligence (AI) usage in the field of toxicologic pathology. In addition to demographic information, some general questions explored AI relative to (1) the current status of adoption across organizations; (2) technical and methodological aspects; (3) perceived business value and finally; and (4) roadblocks and perspectives. AI has become increasingly established in toxicologic pathology with most pathologists being supportive of its development despite some areas of uncertainty. A salient feature consisted of the variability of AI awareness and adoption among the responders, as the spectrum extended from pathologists having developed familiarity and technical skills in AI, to colleagues who had no interest in AI as a tool in toxicologic pathology. Despite a general enthusiasm for these techniques, the overall understanding and trust in AI algorithms as well as their added value in toxicologic pathology were generally low, suggesting room for the need for increased awareness and education. This survey will serve as a basis to evaluate the evolution of AI penetration and acceptance in this domain.

Deep learning generates synthetic cancer histology for explainability and education

Artificial intelligence methods including deep neural networks (DNN) can provide rapid molecular classification of tumors from routine histology with accuracy that matches or exceeds human pathologists. Discerning how neural networks make their predictions remains a significant challenge, but explainability tools help provide insights into what models have learned when corresponding histologic features are poorly defined. Here, we present a method for improving explainability of DNN models using synthetic histology generated by a conditional generative adversarial network (cGAN). We show that cGANs generate high-quality synthetic histology images that can be leveraged for explaining DNN models trained to classify molecularly-subtyped tumors, exposing histologic features associated with molecular state. Fine-tuning synthetic histology through class and layer blending illustrates nuanced morphologic differences between tumor subtypes. Finally, we demonstrate the use of synthetic histology for augmenting pathologist-in-training education, showing that these intuitive visualizations can reinforce and improve understanding of histologic manifestations of tumor biology.

Generalizable biomarker prediction from cancer pathology slides with self-supervised deep learning: A retrospective multi-centric study

Deep learning (DL) can predict microsatellite instability (MSI) from routine histopathology slides of colorectal cancer (CRC). However, it is unclear whether DL can also predict other biomarkers with high performance and whether DL predictions generalize to external patient populations. Here, we acquire CRC tissue samples from two large multi-centric studies. We systematically compare six different state-of-the-art DL architectures to predict biomarkers from pathology slides, including MSI and mutations in BRAF, KRAS, NRAS, and PIK3CA. Using a large external validation cohort to provide a realistic evaluation setting, we show that models using self-supervised, attention-based multiple-instance learning consistently outperform previous approaches while offering explainable visualizations of the indicative regions and morphologies. While the prediction of MSI and BRAF mutations reaches a clinical-grade performance, mutation prediction of PIK3CA, KRAS, and NRAS was clinically insufficient.

Direct prediction of genetic aberrations from pathology images in gastric cancer with swarm learning

BACKGROUND

Computational pathology uses deep learning (DL) to extract biomarkers from routine pathology slides. Large multicentric datasets improve performance, but such datasets are scarce for gastric cancer. This limitation could be overcome by Swarm Learning (SL).

METHODS

Here, we report the results of a multicentric retrospective study of SL for prediction of molecular biomarkers in gastric cancer. We collected tissue samples with known microsatellite instability (MSI) and Epstein-Barr Virus (EBV) status from four patient cohorts from Switzerland, Germany, the UK and the USA, storing each dataset on a physically separate computer.

RESULTS

On an external validation cohort, the SL-based classifier reached an area under the receiver operating curve (AUROC) of 0.8092 (± 0.0132) for MSI prediction and 0.8372 (± 0.0179) for EBV prediction. The centralized model, which was trained on all datasets on a single computer, reached a similar performance.

CONCLUSIONS

Our findings demonstrate the feasibility of SL-based molecular biomarkers in gastric cancer. In the future, SL could be used for collaborative training and, thus, improve the performance of these biomarkers. This may ultimately result in clinical-grade performance and generalizability.

Computational pathology in 2030: a Delphi study forecasting the role of AI in pathology within the next decade

BACKGROUND

Artificial intelligence (AI) is rapidly fuelling a fundamental transformation in the practice of pathology. However, clinical integration remains challenging, with no AI algorithms to date in routine adoption within typical anatomic pathology (AP) laboratories. This survey gathered current expert perspectives and expectations regarding the role of AI in AP from those with first-hand computational pathology and AI experience.

METHODS

Perspectives were solicited using the Delphi method from 24 subject matter experts between December 2020 and February 2021 regarding the anticipated role of AI in pathology by the year 2030. The study consisted of three consecutive rounds: 1) an open-ended, free response questionnaire generating a list of survey items; 2) a Likert-scale survey scored by experts and analysed for consensus; and 3) a repeat survey of items not reaching consensus to obtain further expert consensus.

FINDINGS

Consensus opinions were reached on 141 of 180 survey items (78.3%). Experts agreed that AI would be routinely and impactfully used within AP laboratory and pathologist clinical workflows by 2030. High consensus was reached on 100 items across nine categories encompassing the impact of AI on (1) pathology key performance indicators (KPIs) and (2) the pathology workforce and specific tasks performed by (3) pathologists and (4) AP lab technicians, as well as (5) specific AI applications and their likelihood of routine use by 2030, (6) AI's role in integrated diagnostics, (7) pathology tasks likely to be fully automated using AI, and (8) regulatory/legal and (9) ethical aspects of AI integration in pathology.

INTERPRETATION

This systematic consensus study details the expected short-to-mid-term impact of AI on pathology practice. These findings provide timely and relevant information regarding future care delivery in pathology and raise key practical, ethical, and legal challenges that must be addressed prior to AI's successful clinical implementation.

FUNDING

No specific funding was provided for this study.

Adversarial attacks and adversarial robustness in computational pathology

Artificial Intelligence (AI) can support diagnostic workflows in oncology by aiding diagnosis and providing biomarkers directly from routine pathology slides. However, AI applications are vulnerable to adversarial attacks. Hence, it is essential to quantify and mitigate this risk before widespread clinical use. Here, we show that convolutional neural networks (CNNs) are highly susceptible to white- and black-box adversarial attacks in clinically relevant weakly-supervised classification tasks. Adversarially robust training and dual batch normalization (DBN) are possible mitigation strategies but require precise knowledge of the type of attack used in the inference. We demonstrate that vision transformers (ViTs) perform equally well compared to CNNs at baseline, but are orders of magnitude more robust to white- and black-box attacks. At a mechanistic level, we show that this is associated with a more robust latent representation of clinically relevant categories in ViTs compared to CNNs. Our results are in line with previous theoretical studies and provide empirical evidence that ViTs are robust learners in computational pathology. This implies that large-scale rollout of AI models in computational pathology should rely on ViTs rather than CNN-based classifiers to provide inherent protection against perturbation of the input data, especially adversarial attacks.

Artificial Intelligence can support diagnostic workflows in oncology, but they are vulnerable to adversarial attacks. Here, the authors show that convolutional neural networks are highly susceptible to white- and black-box adversarial attacks in clinically relevant classification tasks.

Artificial intelligence in histopathology: enhancing cancer research and clinical oncology

Artificial intelligence (AI) methods have multiplied our capabilities to extract quantitative information from digital histopathology images. AI is expected to reduce workload for human experts, improve the objectivity and consistency of pathology reports, and have a clinical impact by extracting hidden information from routinely available data. Here, we describe how AI can be used to predict cancer outcome, treatment response, genetic alterations and gene expression from digitized histopathology slides. We summarize the underlying technologies and emerging approaches, noting limitations, including the need for data sharing and standards. Finally, we discuss the broader implications of AI in cancer research and oncology.