Image-based consensus molecular subtype (imCMS) classification of colorectal cancer using deep learning

Experiences of running a stratified medicine adaptive platform trial: Challenges and lessons learned from 10 years of the FOCUS4 trial in metastatic colorectal cancer

BACKGROUND

Complex innovative design trials are becoming increasingly common and offer potential for improving patient outcomes in a faster time frame. FOCUS4 was the first molecularly stratified trial in metastatic colorectal cancer and it remains one of the first umbrella trial designs to be launched globally. Here, we aim to describe lessons learned from delivery of the trial over the last 10 years.

METHODS

FOCUS4 was a Phase II/III molecularly stratified umbrella trial testing the safety and efficacy of targeted therapies in metastatic colorectal cancer. It used adaptive statistical methodology to decide which sub-trial should close early, and new therapies were added as protocol amendments. Patients with newly diagnosed metastatic colorectal cancer were registered, and central laboratory testing was used to stratify their tumour into molecular subtypes. Following 16 weeks of first-line therapy, patients with stable or responding disease were eligible for randomisation into either a molecularly stratified sub-trial (FOCUS4-B, C or D) or non-stratified FOCUS4-N. The primary outcome for all studies was progression-free survival comparing the intervention with active monitoring/placebo. At the close of the trial, feedback was elicited from all investigators through surveys and interviews and consolidated into a series of recommendations and lessons learned for the delivery of similar future trials.

RESULTS

Between January 2014 and October 2020, 1434 patients were registered from 88 UK hospitals. Of the 20 drug combinations that were explored for inclusion in the platform trial, three molecularly targeted sub-trials were activated: FOCUS4-D (February 2014-March 2016) evaluated AZD8931 in the BRAF-PIK3CA-RAS wildtype subgroup; FOCUS4-B (February 2016-July 2018) evaluated aspirin in the PIK3CA mutant subgroup and FOCUS4-C (June 2017-October 2020) evaluated adavosertib in the RAS+TP53 double mutant subgroup. FOCUS4-N was active throughout and evaluated capecitabine monotherapy versus a treatment break. A total of 361 (25%) registered patients were randomised into a sub-trial. Feedback on the experiences of delivery of FOCUS4 could be grouped into three main areas of challenge: funding/infrastructure, biomarker testing procedures and trial design efficiencies within which 20 recommendations are summarised.

CONCLUSION

Adaptive stratified medicine platform studies are feasible in common cancers but present challenges. Our stakeholder feedback has helped to inform how these trial designs can succeed and answer multiple questions efficiently, providing resource is adequate.

Deep Learning on Histopathological Images for Colorectal Cancer Diagnosis: A Systematic Review

Colorectal cancer (CRC) is the second most common cancer in women and the third most common in men, with an increasing incidence. Pathology diagnosis complemented with prognostic and predictive biomarker information is the first step for personalized treatment. The increased diagnostic load in the pathology laboratory, combined with the reported intra- and inter-variability in the assessment of biomarkers, has prompted the quest for reliable machine-based methods to be incorporated into the routine practice. Recently, Artificial Intelligence (AI) has made significant progress in the medical field, showing potential for clinical applications. Herein, we aim to systematically review the current research on AI in CRC image analysis. In histopathology, algorithms based on Deep Learning (DL) have the potential to assist in diagnosis, predict clinically relevant molecular phenotypes and microsatellite instability, identify histological features related to prognosis and correlated to metastasis, and assess the specific components of the tumor microenvironment.

Artificial intelligence to identify genetic alterations in conventional histopathology

Precision oncology relies on the identification of targetable molecular alterations in tumor tissues. In many tumor types, a limited set of molecular tests is currently part of standard diagnostic workflows. However, universal testing for all targetable alterations, especially rare ones, is limited by the cost and availability of molecular assays. From 2017 to 2021, multiple studies have shown that artificial intelligence (AI) methods can predict the probability of specific genetic alterations directly from conventional hematoxylin and eosin (H&E) tissue slides. Although these methods are currently less accurate than gold-standard testing (e.g. immunohistochemistry, polymerase chain reaction or next-generation sequencing), they could be used as pre-screening tools to reduce the workload of genetic analyses. In this systematic literature review, we summarize the state of the art in predicting molecular alterations from H&E using AI. We found that AI methods perform reasonably well across multiple tumor types, although few algorithms have been broadly validated. In addition, we found that genetic alterations in FGFR, IDH, PIK3CA, BRAF, TP53 and DNA repair pathways are predictable from H&E in multiple tumor types, while many other genetic alterations have rarely been investigated or were only poorly predictable. Finally, we discuss the next steps for the implementation of AI-based surrogate tests in diagnostic workflows. This article is protected by copyright. All rights reserved.

Automated quality assessment of large digitised histology cohorts by artificial intelligence

Research using whole slide images (WSIs) of histopathology slides has increased exponentially over recent years. Glass slides from retrospective cohorts, some with patient follow-up data are digitised for the development and validation of artificial intelligence (AI) tools. Such resources, therefore, become very important, with the need to ensure that their quality is of the standard necessary for downstream AI development. However, manual quality control of large cohorts of WSIs by visual assessment is unfeasible, and whilst quality control AI algorithms exist, these focus on bespoke aspects of image quality, e.g. focus, or use traditional machine-learning methods, which are unable to classify the range of potential image artefacts that should be considered. In this study, we have trained and validated a multi-task deep neural network to automate the process of quality control of a large retrospective cohort of prostate cases from which glass slides have been scanned several years after production, to determine both the usability of the images at the diagnostic level (considered in this study to be the minimal standard for research) and the common image artefacts present. Using a two-layer approach, quality overlays of WSIs were generated from a quality assessment (QA) undertaken at patch-level at \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$5\times$$\end{document}5× magnification. From these quality overlays the slide-level quality scores were predicted and then compared to those generated by three specialist urological pathologists, with a Pearson correlation of 0.89 for overall ‘usability’ (at a diagnostic level), and 0.87 and 0.82 for focus and H&E staining quality scores respectively. To demonstrate its wider potential utility, we subsequently applied our QA pipeline to the TCGA prostate cancer cohort and to a colorectal cancer cohort, for comparison. Our model, designated as PathProfiler, indicates comparable predicted usability of images from the cohorts assessed (86–90% of WSIs predicted to be usable), and perhaps more significantly is able to predict WSIs that could benefit from an intervention such as re-scanning or re-staining for quality improvement. We have shown in this study that AI can be used to automate the process of quality control of large retrospective WSI cohorts to maximise their utility for research.

A Quantitative Comparison between Shannon and Tsallis–Havrda–Charvat Entropies Applied to Cancer Outcome Prediction

In this paper, we propose to quantitatively compare loss functions based on parameterized Tsallis–Havrda–Charvat entropy and classical Shannon entropy for the training of a deep network in the case of small datasets which are usually encountered in medical applications. Shannon cross-entropy is widely used as a loss function for most neural networks applied to the segmentation, classification and detection of images. Shannon entropy is a particular case of Tsallis–Havrda–Charvat entropy. In this work, we compare these two entropies through a medical application for predicting recurrence in patients with head–neck and lung cancers after treatment. Based on both CT images and patient information, a multitask deep neural network is proposed to perform a recurrence prediction task using cross-entropy as a loss function and an image reconstruction task. Tsallis–Havrda–Charvat cross-entropy is a parameterized cross-entropy with the parameter α. Shannon entropy is a particular case of Tsallis–Havrda–Charvat entropy for α=1. The influence of this parameter on the final prediction results is studied. In this paper, the experiments are conducted on two datasets including in total 580 patients, of whom 434 suffered from head–neck cancers and 146 from lung cancers. The results show that Tsallis–Havrda–Charvat entropy can achieve better performance in terms of prediction accuracy with some values of α.

Fibroblast activation protein identifies Consensus Molecular Subtype 4 in colorectal cancer and allows its detection by 68Ga-FAPI-PET imaging

Background

In colorectal cancer (CRC), the consensus molecular subtype 4 (CMS4) is associated with therapy resistance and poor prognosis. Clinical diagnosis of CMS4 is hampered by locoregional and temporal variables influencing CMS classification. Diagnostic tools that comprehensively detect CMS4 are therefore urgently needed.

Methods

To identify targets for molecular CMS4 imaging, RNA sequencing data of 3232 primary CRC patients were explored. Heterogeneity of marker expression in relation to CMS4 status was assessed by analysing 3–5 tumour regions and 91.103 single-tumour cells (7 and 29 tumours, respectively). Candidate marker expression was validated in CMS4 peritoneal metastases (PM; n = 59). Molecular imaging was performed using the ⁶⁸Ga-DOTA-FAPI-46 PET tracer.

Results

Fibroblast activation protein (FAP) mRNA identified CMS4 with very high sensitivity and specificity (AUROC > 0.91), and was associated with significantly shorter relapse-free survival (P = 0.0038). Heterogeneous expression of FAP among and within tumour lesions correlated with CMS4 heterogeneity (AUROC = 1.00). FAP expression was homogeneously high in PM, a near-homogeneous CMS4 entity. FAPI-PET identified focal and diffuse PM that were missed using conventional imaging. Extra-peritoneal metastases displayed extensive heterogeneity of tracer uptake.

Conclusion

FAP expression identifies CMS4 CRC. FAPI-PET may have value in the comprehensive detection of CMS4 tumours in CRC. This is especially relevant in patients with PM, for whom effective imaging tools are currently lacking.

Deep Learning With Radiomics for Disease Diagnosis and Treatment: Challenges and Potential

The high-throughput extraction of quantitative imaging features from medical images for the purpose of radiomic analysis, i.e., radiomics in a broad sense, is a rapidly developing and emerging research field that has been attracting increasing interest, particularly in multimodality and multi-omics studies. In this context, the quantitative analysis of multidimensional data plays an essential role in assessing the spatio-temporal characteristics of different tissues and organs and their microenvironment. Herein, recent developments in this method, including manually defined features, data acquisition and preprocessing, lesion segmentation, feature extraction, feature selection and dimension reduction, statistical analysis, and model construction, are reviewed. In addition, deep learning-based techniques for automatic segmentation and radiomic analysis are being analyzed to address limitations such as rigorous workflow, manual/semi-automatic lesion annotation, and inadequate feature criteria, and multicenter validation. Furthermore, a summary of the current state-of-the-art applications of this technology in disease diagnosis, treatment response, and prognosis prediction from the perspective of radiology images, multimodality images, histopathology images, and three-dimensional dose distribution data, particularly in oncology, is presented. The potential and value of radiomics in diagnostic and therapeutic strategies are also further analyzed, and for the first time, the advances and challenges associated with dosiomics in radiotherapy are summarized, highlighting the latest progress in radiomics. Finally, a robust framework for radiomic analysis is presented and challenges and recommendations for future development are discussed, including but not limited to the factors that affect model stability (medical big data and multitype data and expert knowledge in medical), limitations of data-driven processes (reproducibility and interpretability of studies, different treatment alternatives for various institutions, and prospective researches and clinical trials), and thoughts on future directions (the capability to achieve clinical applications and open platform for radiomics analysis).

From Mouse to Human: Cellular Morphometric Subtype Learned From Mouse Mammary Tumors Provides Prognostic Value in Human Breast Cancer

Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan-Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care.

Deep Learning based on hematoxylin-eosin staining outperforms immunohistochemistry in predicting molecular subtypes of gastric adenocarcinoma

In gastric cancer (GC), there are four molecular subclasses that indicate whether patients respond to chemotherapy or immunotherapy, according to the TCGA. In clinical practice, however, not every patient undergoes molecular testing. Many laboratories have used well-implemented in situ techniques (IHC and EBER-ISH) to determine the subclasses in their cohorts. Although multiple stains are used, we show that a staining approach is unable to correctly discriminate all subclasses. As an alternative, we trained an ensemble convolutional neuronal network using bagging that can predict the molecular subclass directly from hematoxylin-eosin histology. We also identified patients with predicted intra-tumoral heterogeneity or with features from multiple subclasses, which challenges the postulated TCGA-based decision tree for GC subtyping. In the future, Deep Learning may enable targeted testing for molecular subtypes and targeted therapy for a broader group of GC patients. This article is protected by copyright. All rights reserved.

Morphology-Based Deep Learning Approach for Predicting Osteogenic Differentiation

Early, high-throughput, and accurate recognition of osteogenic differentiation of stem cells is urgently required in stem cell therapy, tissue engineering, and regenerative medicine. In this study, we established an automatic deep learning algorithm, i.e., osteogenic convolutional neural network (OCNN), to quantitatively measure the osteogenic differentiation of rat bone marrow mesenchymal stem cells (rBMSCs). rBMSCs stained with F-actin and DAPI during early differentiation (day 0, 1, 4, and 7) were captured using laser confocal scanning microscopy to train OCNN. As a result, OCNN successfully distinguished differentiated cells at a very early stage (24 h) with a high area under the curve (AUC) (0.94 ± 0.04) and correlated with conventional biochemical markers. Meanwhile, OCNN exhibited better prediction performance compared with the single morphological parameters and support vector machine. Furthermore, OCNN successfully predicted the dose-dependent effects of small-molecule osteogenic drugs and a cytokine. OCNN-based online learning models can further recognize the osteogenic differentiation of rBMSCs cultured on several material surfaces. Hence, this study initially demonstrated the foreground of OCNN in osteogenic drug and biomaterial screening for next-generation tissue engineering and stem cell research.

Development of an Aging-Related Gene Signature for Predicting Prognosis, Immunotherapy, and Chemotherapy Benefits in Rectal Cancer

Objective: Aging is the major risk factor for human cancers, including rectal cancer. Targeting the aging process provides broad-spectrum protection against cancers. Here, we investigate the clinical implications of aging-related genes in rectal cancer.

Methods: Dysregulated aging-related genes were screened in rectal cancer from TCGA project. A LASSO prognostic model was conducted, and the predictive performance was evaluated and externally verified in the GEO data set. Associations of the model with tumor-infiltrating immune cells, immune and stromal score, HLA and immune checkpoints, and response to chemotherapeutic agents were analyzed across rectal cancer. Biological processes underlying the model were investigated through GSVA and GSEA methods. Doxorubicin (DOX)-induced or replicative senescent stromal cells were constructed, and AGTR1 was silenced in HUVECs. After coculture with conditioned medium of HUVECs, rectal cancer cell growth and invasion were investigated.

Results: An aging-related model was established, consisting of KL, BRCA1, CLU, and AGTR1, which can stratify high- and low-risk patients in terms of overall survival, disease-free survival, and progression-free interval. ROC and Cox regression analyses confirmed that the model was a robust and independent predictor. Furthermore, it was in relation to tumor immunity and stromal activation as well as predicted the responses to gemcitabine and sunitinib. AGTR1 knockdown ameliorated stromal cell senescence and suppressed senescent stromal cell-triggered rectal cancer progression.

Conclusion: Our findings suggest that the aging-related gene signature was in relation to tumor immunity and stromal activation in rectal cancer, which might predict survival outcomes and immuno- and chemotherapy benefits.

Identification of Tissue Types and Gene Mutations From Histopathology Images for Advancing Colorectal Cancer Biology

Objective: Colorectal cancer (CRC) patients respond differently to treatments and are sub-classified by different approaches. We evaluated a deep learning model, which adopted endoscopic knowledge learnt from AI-doscopist, to characterise CRC patients by histopathological features. Results: Data of 461 patients were collected from TCGA-COAD database. The proposed framework was able to 1) differentiate tumour from normal tissues with an Area Under Receiver Operating Characteristic curve (AUROC) of 0.97; 2) identify certain gene mutations (MYH9, TP53) with an AUROC > 0.75; 3) classify CMS2 and CMS4 better than the other subtypes; and 4) demonstrate the generalizability of predicting KRAS mutants in an external cohort. Conclusions: Artificial intelligent can be used for on-site patient classification. Although KRAS mutants were commonly associated with therapeutic resistance and poor prognosis, subjects with predicted KRAS mutants in this study have a higher survival rate in 30 months after diagnoses.

The seen and the unseen: Molecular classification and image based-analysis of gastrointestinal cancers

Gastrointestinal cancers account for 22.5% of cancer related deaths worldwide and represent circa 20% of all cancers. In the last decades, we have witnessed a shift from histology-based to molecular-based classifications using genomic, epigenomic, and transcriptomic data. The molecular based classification revealed new prognostic markers and may aid the therapy selection. Because of the high-costs to perform a molecular classification, in recent years immunohistochemistry-based surrogate classification were developed which permit the stratification of patients, and in parallel multiple groups developed hematoxylin and eosin whole slide image analysis for sub-classifying these entities. Hence, we are witnessing a return to an image-based classification with the purpose to infer hidden information from routine histology images that would permit to detect the patients that respond to specific therapies and would be able to predict their outcome. In this review paper, we will discuss the current histological, molecular, and immunohistochemical classifications of the most common gastrointestinal cancers, gastric adenocarcinoma, and colorectal adenocarcinoma, and will present key aspects for developing a new artificial intelligence aided image-based classification of these malignancies.

Image-based assessment of extracellular mucin-to-tumor area predicts consensus molecular subtypes (CMS) in colorectal cancer

The backbone of all colorectal cancer classifications including the consensus molecular subtypes (CMS) highlights microsatellite instability (MSI) as a key molecular pathway. Although mucinous histology (generally defined as >50% extracellular mucin-to-tumor area) is a "typical" feature of MSI, it is not limited to this subgroup. Here, we investigate the association of CMS classification and mucin-to-tumor area quantified using a deep learning algorithm, and  the expression of specific mucins in predicting CMS groups and clinical outcome. A weakly supervised segmentation method was developed to quantify extracellular mucin-to-tumor area in H&E images. Performance was compared to two pathologists' scores, then applied to two cohorts: (1) TCGA (n = 871 slides/412 patients) used for mucin-CMS group correlation and (2) Bern (n = 775 slides/517 patients) for histopathological correlations and next-generation Tissue Microarray construction. TCGA and CPTAC (n = 85 patients) were used to further validate mucin detection and CMS classification by gene and protein expression analysis for MUC2, MUC4, MUC5AC and MUC5B. An excellent inter-observer agreement between pathologists' scores and the algorithm was obtained (ICC = 0.92). In TCGA, mucinous tumors were predominantly CMS1 (25.7%), CMS3 (24.6%) and CMS4 (16.2%). Average mucin in CMS2 was 1.8%, indicating negligible amounts. RNA and protein expression of MUC2, MUC4, MUC5AC and MUC5B were low-to-absent in CMS2. MUC5AC protein expression correlated with aggressive tumor features (e.g., distant metastases (p = 0.0334), BRAF mutation (p < 0.0001), mismatch repair-deficiency (p < 0.0001), and unfavorable 5-year overall survival (44% versus 65% for positive/negative staining). MUC2 expression showed the opposite trend, correlating with less lymphatic (p = 0.0096) and venous vessel invasion (p = 0.0023), no impact on survival.The absence of mucin-expressing tumors in CMS2 provides an important phenotype-genotype correlation. Together with MSI, mucinous histology may help predict CMS classification using only histopathology and should be considered in future image classifiers of molecular subtypes.

Artificial intelligence in pathological evaluation of gastrointestinal cancers

The integration of artificial intelligence (AI) has shown promising benefits in many fields of diagnostic histopathology, including for gastrointestinal cancers (GCs), such as tumor identification, classification, and prognosis prediction. In parallel, recent evidence suggests that AI may help reduce the workload in gastrointestinal pathology by automatically detecting tumor tissues and evaluating prognostic parameters. In addition, AI seems to be an attractive tool for biomarker/genetic alteration prediction in GC, as it can contain a massive amount of information from visual data that is complex and partially understandable by pathologists. From this point of view, it is suggested that advances in AI could lead to revolutionary changes in many fields of pathology. Unfortunately, these findings do not exclude the possibility that there are still many hurdles to overcome before AI applications can be safely and effectively applied in actual pathology practice. These include a broad spectrum of challenges from needs identification to cost-effectiveness. Therefore, unlike other disciplines of medicine, no histopathology-based AI application, including in GC, has ever been approved either by a regulatory authority or approved for public reimbursement. The purpose of this review is to present data related to the applications of AI in pathology practice in GC and present the challenges that need to be overcome for their implementation.

Comprehensive Imaging Characterization of Colorectal Liver Metastases

Colorectal liver metastases (CRLM) have heterogenous histopathological and immunohistochemical phenotypes, which are associated with variable responses to treatment and outcomes. However, this information is usually only available after resection, and therefore of limited value in treatment planning. Improved techniques for in vivo disease assessment, which can characterise the variable tumour biology, would support further personalization of management strategies. Advanced imaging of CRLM including multiparametric MRI and functional imaging techniques have the potential to provide clinically-actionable phenotypic characterisation. This includes assessment of the tumour-liver interface, internal tumour components and treatment response. Advanced analysis techniques, including radiomics and machine learning now have a growing role in assessment of imaging, providing high-dimensional imaging feature extraction which can be linked to clinical relevant tumour phenotypes, such as a the Consensus Molecular Subtypes (CMS). In this review, we outline how imaging techniques could reproducibly characterize the histopathological features of CRLM, with several matched imaging and histology examples to illustrate these features, and discuss the oncological relevance of these features. Finally, we discuss the future challenges and opportunities of CRLM imaging, with a focus on the potential value of advanced analytics including radiomics and artificial intelligence, to help inform future research in this rapidly moving field.

Synchronous biliary gallstones and colorectal cancer: A single center analysis

Gallstones and colorectal cancer (CRC) are two common disorders that may develop simultaneously. In such situations, there is a significant chance of missing one of the conditions due to the primary clinical presentation. Late detection, diagnosis and treatment can be especially problematic in the case of unrecognized CRC. In the present study, the medical charts were retrospectively reviewed for all consecutive patients who were treated in the Second Department of Surgery, University Emergency Hospital Bucharest (Romania) between February 2015 and December 2017 following a diagnosis of CRC and/or biliary stones. There were 203 patients with CRC, 433 with biliary gallstones and 19 patients with both conditions. There were 125 men (61.6%) in the CRC group and 138 men (31.9%) in the gallstone group. The average age was 54.1±15.9 years in the gallstone group and 66.1±11.6 years in the CRC group. Obesity was observed in 96 patients (22.2%) with gallstones and in 14 (6.9%) patients in the CRC group. In the CRC group, 80 patients had medical comorbidities (39.4%), while in the gallstone group 126 patients (29.1%) had medical comorbidities. Bivariate analysis comparing gallstone only vs. gallstone and CRC identified age (P=0.001), male sex (P=0.001) and thyroid disease (P=0.001) as significant factors associated with synchronous diagnosis. The multivariable logistic regression of factors predicting CRC in patients with gallstones identified age (OR, 1.06; 95% CI, 1.023-1.105; P=0.002) and thyroid diseases (OR, 11.15; 95% CI, 2.532-49.06; P=0.001) as independent factors. There were significant differences regarding the location of the tumor between the CRC-only group and the gallstone and CRC group (P=0.001): Rectum (39.7 vs. 5.3%), left colon (26.6 vs. 21.1%), transverse colon (13 vs. 26.3%) and right colon (20.7 vs. 47.4%). The study concluded that, in patients with gallstones, age and thyroid conditions were significantly associated with CRC. Patients with a synchronous diagnosis of gallstones and CRC had significantly more right-sided CRC compared with regular CRC.

Pyroptosis-Mediated Molecular Subtypes and Tumor Microenvironment Infiltration Characterization in Colon Cancer

The role of pyroptosis, which is also a kind of cell-intrinsic death mechanism, in tumorigenesis and cancer progression has been revolutionized. However, the expression of pyroptosis-related genes (PYGs) in colon cancer (CC) and their prognostic value remain unclear. In this study, we comprehensively identified two PYG-mediated molecular subtypes with a distinct tumor microenvironment (TME) in 1,415 CC samples, which were based on 10 PYGs. The six-gene signature (pyroptosis score, PY-score) was constructed to quantify the molecular patterns of individual tumors using a least absolute shrinkage and selection operator (LASSO)–Cox regression model through the differentially expressed genes between the two molecular subtypes. Significant infiltration of activated immune cells (such as M1 macrophages and cytotoxic T cells) was observed in the low PY-score group, while naive and suppressive immune cells (such as naive CD8⁺ T cells and M2 macrophages) dominated in the high PY-score group. CC patients in the low PY-score group showed not only significant survival advantage but also sensitivity to immune checkpoint inhibitor treatment, anti-epidermal growth factor receptor (EGFR) therapy, and chemotherapy. Overall, this work revealed that the PYGs played a vital role in the formation of heterogeneity in the TME. The analysis of the PYG-mediated molecular patterns helps in understanding the characterization of TME infiltration and provides insights into more effective therapeutic strategies.

Development and validation of a weakly supervised deep learning framework to predict the status of molecular pathways and key mutations in colorectal cancer from routine histology images: a retrospective study

BACKGROUND

Determining the status of molecular pathways and key mutations in colorectal cancer is crucial for optimal therapeutic decision making. We therefore aimed to develop a novel deep learning pipeline to predict the status of key molecular pathways and mutations from whole-slide images of haematoxylin and eosin-stained colorectal cancer slides as an alternative to current tests.

METHODS

In this retrospective study, we used 502 diagnostic slides of primary colorectal tumours from 499 patients in The Cancer Genome Atlas colon and rectal cancer (TCGA-CRC-DX) cohort and developed a weakly supervised deep learning framework involving three separate convolutional neural network models. Whole-slide images were divided into equally sized tiles and model 1 (ResNet18) extracted tumour tiles from non-tumour tiles. These tumour tiles were inputted into model 2 (adapted ResNet34), trained by iterative draw and rank sampling to calculate a prediction score for each tile that represented the likelihood of a tile belonging to the molecular labels of high mutation density (vs low mutation density), microsatellite instability (vs microsatellite stability), chromosomal instability (vs genomic stability), CpG island methylator phenotype (CIMP)-high (vs CIMP-low), BRAFmut (vs BRAFWT), TP53mut (vs TP53WT), and KRASWT (vs KRASmut). These scores were used to identify the top-ranked titles from each slide, and model 3 (HoVer-Net) segmented and classified the different types of cell nuclei in these tiles. We calculated the area under the convex hull of the receiver operating characteristic curve (AUROC) as a model performance measure and compared our results with those of previously published methods.

FINDINGS

Our iterative draw and rank sampling method yielded mean AUROCs for the prediction of hypermutation (0·81 [SD 0·03] vs 0·71), microsatellite instability (0·86 [0·04] vs 0·74), chromosomal instability (0·83 [0·02] vs 0·73), BRAFmut (0·79 [0·01] vs 0·66), and TP53mut (0·73 [0·02] vs 0·64) in the TCGA-CRC-DX cohort that were higher than those from previously published methods, and an AUROC for KRASmut that was similar to previously reported methods (0·60 [SD 0·04] vs 0·60). Mean AUROC for predicting CIMP-high status was 0·79 (SD 0·05). We found high proportions of tumour-infiltrating lymphocytes and necrotic tumour cells to be associated with microsatellite instability, and high proportions of tumour-infiltrating lymphocytes and a low proportion of necrotic tumour cells to be associated with hypermutation.

INTERPRETATION

After large-scale validation, our proposed algorithm for predicting clinically important mutations and molecular pathways, such as microsatellite instability, in colorectal cancer could be used to stratify patients for targeted therapies with potentially lower costs and quicker turnaround times than sequencing-based or immunohistochemistry-based approaches.

FUNDING

The UK Medical Research Council.

The therapeutic and prognostic implications of immunobiology in colorectal cancer: a review

Colorectal cancer represents the second leading cause of cancer-related death worldwide. The therapeutic field of immuno-oncology has rapidly gained momentum, with strikingly promising results observed in clinical practice. Increasing emphasis has been placed on the role of the immune response in tumorigenesis, therapy and predicting prognosis. Enhanced understanding of the dynamic and complex tumour-immune microenvironment has enabled the development of molecularly directed, individualised treatment. Analysis of intra-tumoural lymphocyte infiltration and the dichotomisation of colorectal cancer into microsatellite stable and unstable disease has important therapeutic and prognostic implications, with potential to capitalise further on this data. This review discusses the latest evidence surrounding the tumour biology and immune landscape of colorectal cancer, novel immunotherapies and the interaction of the immune system with each apex of the tripartite of cancer management (oncotherapeutics, radiotherapy and surgery). By utilising the synergy of chemotherapeutic agents and immunotherapies, and identifying prognostic and predictive immunological biomarkers, we may enter an era of unprecedented disease control, survivorship and cure rates.

Deep learning in cancer diagnosis, prognosis and treatment selection

Deep learning is a subdiscipline of artificial intelligence that uses a machine learning technique called artificial neural networks to extract patterns and make predictions from large data sets. The increasing adoption of deep learning across healthcare domains together with the availability of highly characterised cancer datasets has accelerated research into the utility of deep learning in the analysis of the complex biology of cancer. While early results are promising, this is a rapidly evolving field with new knowledge emerging in both cancer biology and deep learning. In this review, we provide an overview of emerging deep learning techniques and how they are being applied to oncology. We focus on the deep learning applications for omics data types, including genomic, methylation and transcriptomic data, as well as histopathology-based genomic inference, and provide perspectives on how the different data types can be integrated to develop decision support tools. We provide specific examples of how deep learning may be applied in cancer diagnosis, prognosis and treatment management. We also assess the current limitations and challenges for the application of deep learning in precision oncology, including the lack of phenotypically rich data and the need for more explainable deep learning models. Finally, we conclude with a discussion of how current obstacles can be overcome to enable future clinical utilisation of deep learning.

Gastrointestinal cancer classification and prognostication from histology using deep learning: Systematic review

BACKGROUND

Gastrointestinal cancers account for approximately 20% of all cancer diagnoses and are responsible for 22.5% of cancer deaths worldwide. Artificial intelligence-based diagnostic support systems, in particular convolutional neural network (CNN)-based image analysis tools, have shown great potential in medical computer vision. In this systematic review, we summarise recent studies reporting CNN-based approaches for digital biomarkers for characterization and prognostication of gastrointestinal cancer pathology.

METHODS

Pubmed and Medline were screened for peer-reviewed papers dealing with CNN-based gastrointestinal cancer analyses from histological slides, published between 2015 and 2020.Seven hundred and ninety titles and abstracts were screened, and 58 full-text articles were assessed for eligibility.

RESULTS

Sixteen publications fulfilled our inclusion criteria dealing with tumor or precursor lesion characterization or prognostic and predictive biomarkers: 14 studies on colorectal or rectal cancer, three studies on gastric cancer and none on esophageal cancer. These studies were categorised according to their end-points: polyp characterization, tumor characterization and patient outcome. Regarding the translation into clinical practice, we identified several studies demonstrating generalization of the classifier with external tests and comparisons with pathologists, but none presenting clinical implementation.

CONCLUSIONS

Results of recent studies on CNN-based image analysis in gastrointestinal cancer pathology are promising, but studies were conducted in observational and retrospective settings. Large-scale trials are needed to assess performance and predict clinical usefulness. Furthermore, large-scale trials are required for approval of CNN-based prediction models as medical devices.

Metastatic heterogeneity of the consensus molecular subtypes of colorectal cancer

Gene expression-based subtypes of colorectal cancer have clinical relevance, but the representativeness of primary tumors and the consensus molecular subtypes (CMS) for metastatic cancers is not well known. We investigated the metastatic heterogeneity of CMS. The best approach to subtype translation was delineated by comparisons of transcriptomic profiles from 317 primary tumors and 295 liver metastases, including multi-metastatic samples from 45 patients and 14 primary-metastasis sets. Associations were validated in an external data set (n = 618). Projection of metastases onto principal components of primary tumors showed that metastases were depleted of CMS1-immune/CMS3-metabolic signals, enriched for CMS4-mesenchymal/stromal signals, and heavily influenced by the microenvironment. The tailored CMS classifier (available in an updated version of the R package CMScaller) therefore implemented an approach to regress out the liver tissue background. The majority of classified metastases were either CMS2 or CMS4. Nonetheless, subtype switching and inter-metastatic CMS heterogeneity were frequent and increased with sampling intensity. Poor-prognostic value of CMS1/3 metastases was consistent in the context of intra-patient tumor heterogeneity.

Status quo and future prospects of artificial neural network from the perspective of gastroenterologists

Artificial neural networks (ANNs) are one of the primary types of artificial intelligence and have been rapidly developed and used in many fields. In recent years, there has been a sharp increase in research concerning ANNs in gastrointestinal (GI) diseases. This state-of-the-art technique exhibits excellent performance in diagnosis, prognostic prediction, and treatment. Competitions between ANNs and GI experts suggest that efficiency and accuracy might be compatible in virtue of technique advancements. However, the shortcomings of ANNs are not negligible and may induce alterations in many aspects of medical practice. In this review, we introduce basic knowledge about ANNs and summarize the current achievements of ANNs in GI diseases from the perspective of gastroenterologists. Existing limitations and future directions are also proposed to optimize ANN’s clinical potential. In consideration of barriers to interdisciplinary knowledge, sophisticated concepts are discussed using plain words and metaphors to make this review more easily understood by medical practitioners and the general public.

Requirements for implementation of artificial intelligence in the practice of gastrointestinal pathology

Tremendous advances in artificial intelligence (AI) in medical image analysis have been achieved in recent years. The integration of AI is expected to cause a revolution in various areas of medicine, including gastrointestinal (GI) pathology. Currently, deep learning algorithms have shown promising benefits in areas of diagnostic histopathology, such as tumor identification, classification, prognosis prediction, and biomarker/genetic alteration prediction. While AI cannot substitute pathologists, carefully constructed AI applications may increase workforce productivity and diagnostic accuracy in pathology practice. Regardless of these promising advances, unlike the areas of radiology or cardiology imaging, no histopathology-based AI application has been approved by a regulatory authority or for public reimbursement. Thus, implying that there are still some obstacles to be overcome before AI applications can be safely and effectively implemented in real-life pathology practice. The challenges have been identified at different stages of the development process, such as needs identification, data curation, model development, validation, regulation, modification of daily workflow, and cost-effectiveness balance. The aim of this review is to present challenges in the process of AI development, validation, and regulation that should be overcome for its implementation in real-life GI pathology practice.

Clinical Value of Consensus Molecular Subtypes in Colorectal Cancer: A Systematic Review and Meta-Analysis

Background

The consensus molecular subtypes (CMSs) of colorectal cancer (CRC) capture tumor heterogeneity at the gene-expression level. Currently, a restricted number of molecular features are used to guide treatment for CRC. We summarize the evidence on the clinical value of the CMSs.

Methods

We systematically identified studies in Medline and Embase that evaluated the prognostic and predictive value of CMSs in CRC patients. A random-effect meta-analysis was performed on prognostic data. Predictive data were summarized.

Results

In local disease, CMS4 tumors were associated with worse overall survival (OS) compared with CMS1 (hazard ratio [HR] = 3.28, 95% confidence interval = 1.27 to 8.47) and CMS2 cancers (HR = 2.60, 95% confidence interval = 1.93 to 3.50). In metastatic disease, CMS1 consistently had worse survival than CMS2-4 (OS HR range = 0.33-0.55; progression-free survival HR range = 0.53-0.89). Adjuvant chemotherapy in stage II and III CRC was most beneficial for OS in CMS2 and CMS3 (HR range = 0.16-0.45) and not effective in CMS4 tumors. In metastatic CMS4 cancers, an irinotecan-based regimen improved outcome compared with oxaliplatin (HR range = 0.31-0.72). The addition of bevacizumab seemed beneficial in CMS1, and anti-epidermal growth factor receptor therapy improved outcome for KRAS wild-type CMS2 patients.

Conclusions

The CMS classification holds clear potential for clinical use in predicting both prognosis and response to systemic therapy, which seems to be independent of the classifier used. Prospective studies are warranted to support implementation of the CMS taxonomy in clinical practice.

State of machine and deep learning in histopathological applications in digestive diseases

Machine learning (ML)- and deep learning (DL)-based imaging modalities have exhibited the capacity to handle extremely high dimensional data for a number of computer vision tasks. While these approaches have been applied to numerous data types, this capacity can be especially leveraged by application on histopathological images, which capture cellular and structural features with their high-resolution, microscopic perspectives. Already, these methodologies have demonstrated promising performance in a variety of applications like disease classification, cancer grading, structure and cellular localizations, and prognostic predictions. A wide range of pathologies requiring histopathological evaluation exist in gastroenterology and hepatology, indicating these as disciplines highly targetable for integration of these technologies. Gastroenterologists have also already been primed to consider the impact of these algorithms, as development of real-time endoscopic video analysis software has been an active and popular field of research. This heightened clinical awareness will likely be important for future integration of these methods and to drive interdisciplinary collaborations on emerging studies. To provide an overview on the application of these methodologies for gastrointestinal and hepatological histopathological slides, this review will discuss general ML and DL concepts, introduce recent and emerging literature using these methods, and cover challenges moving forward to further advance the field.

Deep learning in histopathology: the path to the clinic

Machine learning techniques have great potential to improve medical diagnostics, offering ways to improve accuracy, reproducibility and speed, and to ease workloads for clinicians. In the field of histopathology, deep learning algorithms have been developed that perform similarly to trained pathologists for tasks such as tumor detection and grading. However, despite these promising results, very few algorithms have reached clinical implementation, challenging the balance between hope and hype for these new techniques. This Review provides an overview of the current state of the field, as well as describing the challenges that still need to be addressed before artificial intelligence in histopathology can achieve clinical value.

Artificial intelligence in gastroenterology and hepatology: Status and challenges

Originally proposed by John McCarthy in 1955, artificial intelligence (AI) has achieved a breakthrough and revolutionized the processing methods of clinical medicine with the increasing workloads of medical records and digital images. Doctors are paying attention to AI technologies for various diseases in the fields of gastroenterology and hepatology. This review will illustrate AI technology procedures for medical image analysis, including data processing, model establishment, and model validation. Furthermore, we will summarize AI applications in endoscopy, radiology, and pathology, such as detecting and evaluating lesions, facilitating treatment, and predicting treatment response and prognosis with excellent model performance. The current challenges for AI in clinical application include potential inherent bias in retrospective studies that requires larger samples for validation, ethics and legal concerns, and the incomprehensibility of the output results. Therefore, doctors and researchers should cooperate to address the current challenges and carry out further investigations to develop more accurate AI tools for improved clinical applications.

Development and initial validation of a deep learning algorithm to quantify histological features in colorectal carcinoma including tumour budding/poorly differentiated clusters

AIMS

To develop and validate a deep learning algorithm to quantify a broad spectrum of histological features in colorectal carcinoma.

METHODS AND RESULTS

A deep learning algorithm was trained on haematoxylin and eosin-stained slides from tissue microarrays of colorectal carcinomas (N = 230) to segment colorectal carcinoma digitised images into 13 regions and one object. The segmentation algorithm demonstrated moderate to almost perfect agreement with interpretations by gastrointestinal pathologists, and was applied to an independent test cohort of digitised whole slides of colorectal carcinoma (N = 136). The algorithm correctly classified mucinous and high-grade tumours, and identified significant differences between mismatch repair-proficient and mismatch repair-deficient (MMRD) tumours with regard to mucin, inflammatory stroma, and tumour-infiltrating lymphocytes (TILs). A cutoff of >44.4 TILs per mm² carcinoma gave a sensitivity of 88% and a specificity of 73% in classifying MMRD carcinomas. Algorithm measures of tumour budding (TB) and poorly differentiated clusters (PDCs) outperformed TB grade derived from routine sign-out, and compared favourably with manual counts of TB/PDCs with regard to lymphatic, venous and perineural invasion. Comparable associations were seen between algorithm measures of TB/PDCs and manual counts of TB/PDCs for lymph node metastasis (all P < 0.001); however, stronger correlations were seen between the proportion of positive lymph nodes and algorithm measures of TB/PDCs. Stronger associations were also seen between distant metastasis and algorithm measures of TB/PDCs (P = 0.004) than between distant metastasis and TB (P = 0.04) and TB/PDC counts (P = 0.06).

CONCLUSIONS

Our results highlight the potential of deep learning to identify and quantify a broad spectrum of histological features in colorectal carcinoma.

Integrated approaches for precision oncology in colorectal cancer: The more you know, the better

Colorectal cancer (CRC) is one of the most common human malignancies accounting for approximately 10 % of worldwide cancer incidence and mortality. While early-stage CRC is mainly a preventable and curable disease, metastatic colorectal cancer (mCRC) remains an unmet clinical need. Moreover, about 25 % of CRC cases are diagnosed only at the metastatic stage. Despite the extensive molecular and functional knowledge on this disease, systemic therapy for mCRC still relies on traditional 5-fluorouracil (5-FU)-based chemotherapy regimens. On the other hand, targeted therapies and immunotherapy have shown effectiveness only in a limited subset of patients. For these reasons, there is a growing need to define the molecular and biological landscape of individual patients to implement novel, rationally driven, tailored therapies. In this review, we explore current and emerging approaches for CRC management such as genomic, transcriptomic and metabolomic analysis, the use of liquid biopsies and the implementation of patients' preclinical avatars. In particular, we discuss the contribution of each of these tools in elucidating patient specific features, with the aim of improving our ability in advancing the diagnosis and treatment of colorectal tumors.

The Potential of Artificial Intelligence to Detect Lymphovascular Invasion in Testicular Cancer

Testicular cancer is the most common cancer in men aged from 15 to 34 years. Lymphovascular invasion refers to the presence of tumours within endothelial-lined lymphatic or vascular channels, and has been shown to have prognostic significance in testicular germ cell tumours. In non-seminomatous tumours, lymphovascular invasion is the most powerful prognostic factor for stage 1 disease. For the pathologist, searching multiple slides for lymphovascular invasion can be highly time-consuming. The aim of this retrospective study was to develop and assess an artificial intelligence algorithm that can identify areas suspicious for lymphovascular invasion in histological digital whole slide images. Areas of possible lymphovascular invasion were annotated in a total of 184 whole slide images of haematoxylin and eosin (H&E) stained tissue from 19 patients with testicular germ cell tumours, including a mixture of seminoma and non-seminomatous cases. Following consensus review by specialist uropathologists, we trained a deep learning classifier for automatic segmentation of areas suspicious for lymphovascular invasion. The classifier identified 34 areas within a validation set of 118 whole slide images from 10 patients, each of which was reviewed by three expert pathologists to form a majority consensus. The precision was 0.68 for areas which were considered to be appropriate to flag, and 0.56 for areas considered to be definite lymphovascular invasion. An artificial intelligence tool which highlights areas of possible lymphovascular invasion to reporting pathologists, who then make a final judgement on its presence or absence, has been demonstrated as feasible in this proof-of-concept study. Further development is required before clinical deployment.

Advances in colon cancer research: in vitro and animal models

Modelling human colon cancer has long been the ambition of researchers and oncologists with the aim to better replicate disease progression and treatment response. Advances in our understanding of genetics, stem cell biology, tumour microenvironment and immunology have prepared the groundwork for recent major advances. In the last two years the field has seen the progression of: using patient derived organoids (alone and in co-culture) as predictors of treatment response; molecular stratification of tumours that predict outcome and treatment response; mouse models of metastatic disease; and transplant models that can be used to de-risk clinical trials. We will discuss these advances in this review.

Deep learning in cancer pathology: a new generation of clinical biomarkers

Clinical workflows in oncology rely on predictive and prognostic molecular biomarkers. However, the growing number of these complex biomarkers tends to increase the cost and time for decision-making in routine daily oncology practice; furthermore, biomarkers often require tumour tissue on top of routine diagnostic material. Nevertheless, routinely available tumour tissue contains an abundance of clinically relevant information that is currently not fully exploited. Advances in deep learning (DL), an artificial intelligence (AI) technology, have enabled the extraction of previously hidden information directly from routine histology images of cancer, providing potentially clinically useful information. Here, we outline emerging concepts of how DL can extract biomarkers directly from histology images and summarise studies of basic and advanced image analysis for cancer histology. Basic image analysis tasks include detection, grading and subtyping of tumour tissue in histology images; they are aimed at automating pathology workflows and consequently do not immediately translate into clinical decisions. Exceeding such basic approaches, DL has also been used for advanced image analysis tasks, which have the potential of directly affecting clinical decision-making processes. These advanced approaches include inference of molecular features, prediction of survival and end-to-end prediction of therapy response. Predictions made by such DL systems could simplify and enrich clinical decision-making, but require rigorous external validation in clinical settings.

Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods

Tissue microarray (TMA) core images are a treasure trove for artificial intelligence applications. However, a common problem of TMAs is multiple sectioning, which can change the content of the intended tissue core and requires re-labelling. Here, we investigate different ensemble methods for colorectal tissue classification using high-throughput TMAs. Hematoxylin and Eosin (H&E) core images of 0.6 mm or 1.0 mm diameter from three international cohorts were extracted from 54 digital slides (n = 15,150 cores). After TMA core extraction and color enhancement, five different flows of independent and ensemble deep learning were applied. Training and testing data with 2144 and 13,006 cores included three classes: tumor, normal or "other" tissue. Ground-truth data were collected from 30 ngTMA slides (n = 8689 cores). A test augmentation is applied to reduce the uncertain prediction. Predictive accuracy of the best method, namely Soft Voting Ensemble of one VGG and one CapsNet models was 0.982, 0.947 and 0.939 for normal, "other" and tumor, which outperformed to independent or ensemble learning with one base-estimator. Our high-accuracy algorithm for colorectal tissue classification in high-throughput TMAs is amenable to images from different institutions, core sizes and stain intensity. It helps to reduce error in TMA core evaluations with previously given labels.

Artificial Intelligence for Histology-Based Detection of Microsatellite Instability and Prediction of Response to Immunotherapy in Colorectal Cancer

Microsatellite instability (MSI) is a molecular marker of deficient DNA mismatch repair (dMMR) that is found in approximately 15% of colorectal cancer (CRC) patients. Testing all CRC patients for MSI/dMMR is recommended as screening for Lynch Syndrome and, more recently, to determine eligibility for immune checkpoint inhibitors in advanced disease. However, universal testing for MSI/dMMR has not been uniformly implemented because of cost and resource limitations. Artificial intelligence has been used to predict MSI/dMMR directly from hematoxylin and eosin (H&E) stained tissue slides. We review the emerging data regarding the utility of machine learning for MSI classification, focusing on CRC. We also provide the clinician with an introduction to image analysis with machine learning and convolutional neural networks. Machine learning can predict MSI/dMMR with high accuracy in high quality, curated datasets. Accuracy can be significantly decreased when applied to cohorts with different ethnic and/or clinical characteristics, or different tissue preparation protocols. Research is ongoing to determine the optimal machine learning methods for predicting MSI, which will need to be compared to current clinical practices, including next-generation sequencing. Predicting response to immunotherapy remains an unmet need.

QuPath: The global impact of an open source digital pathology system

QuPath, originally created at the Centre for Cancer Research & Cell Biology at Queen's University Belfast as part of a research programme in digital pathology (DP) funded by Invest Northern Ireland and Cancer Research UK, is arguably the most wildly used image analysis software program in the world. On the back of the explosion of DP and a need to comprehensively visualise and analyse whole slides images (WSI), QuPath was developed to address the many needs associated with tissue based image analysis; these were several fold and, predominantly, translational in nature: from the requirement to visualise images containing billions of pixels from files several GBs in size, to the demand for high-throughput reproducible analysis, which the paradigm of routine visual pathological assessment continues to struggle to deliver. Resultantly, large-scale biomarker quantification must increasingly be augmented with DP. Here we highlight the impact of the open source Quantitative Pathology & Bioimage Analysis DP system since its inception, by discussing the scope of scientific research in which QuPath has been cited, as the system of choice for researchers.

Deep Learning for Whole Slide Image Analysis: An Overview

The widespread adoption of whole slide imaging has increased the demand for effective and efficient gigapixel image analysis. Deep learning is at the forefront of computer vision, showcasing significant improvements over previous methodologies on visual understanding. However, whole slide images have billions of pixels and suffer from high morphological heterogeneity as well as from different types of artifacts. Collectively, these impede the conventional use of deep learning. For the clinical translation of deep learning solutions to become a reality, these challenges need to be addressed. In this paper, we review work on the interdisciplinary attempt of training deep neural networks using whole slide images, and highlight the different ideas underlying these methodologies.