The lightning development of artificial intelligence (AI) has revolutionized healthcare, helping significant improvements in various applications. This paper provides a comprehensive review of foundation models in healthcare, highlighting their transformative potential in areas such as diagnostics, personalized treatment, and operational efficiency. We argue the key capabilities of these models, including their ability to process diverse data types such as medical images, clinical notes, and structured health records. Regardless their assurance, difficulties remain, including data privacy concerns, bias in AI algorithms, and the need for extensive computational resources. Our analysis identifies emerging trends and future directions, emphasizing the importance of ethical AI deployment, improved interoperability over healthcare systems, and the development of more robust, domain-specific models. Future research should focus on enhancing model interpretability, ensuring equitable access, and fostering collaboration between AI developers and healthcare professionals to maximize the advantages of these technologies.

Over 90% of gastrointestinal stromal tumors (GISTs) harbor mutations in KIT or PDGFRA that can predict response to tyrosine kinase inhibitor (TKI) therapies, as recommended by NCCN (National Comprehensive Cancer Network) guidelines. However, gene sequencing for mutation testing is expensive and time-consuming and is susceptible to a variety of preanalytical factors. To overcome the challenges associated with genetic screening by sequencing, in the current study we developed an artificial intelligence-based deep-learning （DL） model that uses convolutional neural networks (CNN) to analyze digitized hematoxylin and eosin staining in tumor histological sections to predict potential response to imatinib or avapritinib treatment in GIST patients. Assessment with an independent testing set showed that our DL  model could predict imatinib sensitivity with an area under the curve (AUC) of 0.902 in case-wise analysis and 0.807 in slide-wise analysis. Case-level AUCs for predicting imatinib-dose-adjustment cases, avapritinib-sensitive cases, and wildtype GISTs were 0.920, 0.958, and 0.776, respectively, while slide-level AUCs for these respective groups were 0.714, 0.922, and 0.886, respectively. Our model showed comparable or better prediction of actual response to TKI than sequencing-based screening (accuracy 0.9286 versus 0.8929; DL model versus sequencing), while predictions of nonresponse to imatinib/avapritinib showed markedly higher accuracy than sequencing (0.7143 versus 0.4286). These results demonstrate the potential of a DL model to improve predictions of treatment response to TKI therapy from histology in GIST patients. © 2025 The Pathological Society of Great Britain and Ireland.

Traditional histopathology, characterized by manual quantifications and assessments, faces challenges such as low-throughput and inter-observer variability that hinder the introduction of precision medicine in pathology diagnostics and research. The advent of digital pathology allowed the introduction of computational pathology, a discipline that leverages computational methods, especially based on deep learning (DL) techniques, to analyze histopathology specimens. A growing body of research shows impressive performances of DL-based models in pathology for a multitude of tasks, such as mutation prediction, large-scale pathomics analyses, or prognosis prediction. New approaches integrate multimodal data sources and increasingly rely on multi-purpose foundation models. This review provides an introductory overview of advancements in computational pathology and discusses their implications for the future of histopathology in research and diagnostics.

The most widespread primary intraocular tumor in adults is called uveal melanoma (UM), if detected early enough, it can be curable. Various methods are available to treat UM, but the most commonly used and effective approach is plaque radiotherapy using Iodine-125 and Ruthenium-106.

The authors performed searches to distinguish relevant studies from 2017 to 2024 by three databases (PubMed, Scopus, and Google Scholar).

Imaging technologies such as ultrasound (US), fundus photography (FP), optical coherent tomography (OCT), fluorescein angiography (FA), and magnetic resonance images (MRI) play a vital role in the diagnosis and prognosis of UM. The present review assessed the power of different image modalities when integrated with artificial intelligence (AI) to diagnose and prognosis of patients affected by UM.

Finally, after reviewing the studies conducted, it was concluded that AI is a developing tool in image analysis and enhances workflows in diagnosis from data and image processing to clinical decisions, improving tailored treatment scenarios, response prediction, and prognostication.

With the rapid advancement of artificial intelligence (AI) technologies, including deep learning, large language models, and neural networks, these methodologies are increasingly being developed and integrated into cancer research. Gastrointestinal tumors are characterized by complexity and heterogeneity, posing significant challenges for early detection, diagnostic accuracy, and the development of personalized treatment strategies. The application of AI in digestive oncology has demonstrated its transformative potential. AI not only alleviates the diagnostic burden on clinicians, but it improves tumor screening sensitivity, specificity, and accuracy. Additionally, AI aids the detection of biomarkers such as microsatellite instability and mismatch repair, supports intraoperative assessments of tumor invasion depth, predicts treatment responses, and facilitates the design of personalized treatment plans to potentially significantly enhance patient outcomes. Moreover, the integration of AI with multiomics analyses and imaging technologies has led to substantial advancements in foundational research on the tumor microenvironment. This review highlights the progress of AI in gastrointestinal oncology over the past 5 years with focus on early tumor screening, diagnosis, molecular marker identification, treatment planning, and prognosis predictions. We also explored the potential of AI to enhance medical imaging analyses to aid tumor detection and characterization as well as its role in automating and refining histopathological assessments.

There is no doubt that advanced artificial intelligence models and high quality data are the keys to success in developing computational pathology tools. Although the overall volume of pathology data keeps increasing, a lack of quality data is a common issue when it comes to a specific task due to several reasons including privacy and ethical issues with patient data. In this work, we propose to exploit knowledge distillation, i.e., utilize the existing model to learn a new, target model, to overcome such issues in computational pathology. Specifically, we employ a student-teacher framework to learn a target model from a pre-trained, teacher model without direct access to source data and distill relevant knowledge via momentum contrastive learning with multi-head attention mechanism, which provides consistent and context-aware feature representations. This enables the target model to assimilate informative representations of the teacher model while seamlessly adapting to the unique nuances of the target data. The proposed method is rigorously evaluated across different scenarios where the teacher model was trained on the same, relevant, and irrelevant classification tasks with the target model. Experimental results demonstrate the accuracy and robustness of our approach in transferring knowledge to different domains and tasks, outperforming other related methods. Moreover, the results provide a guideline on the learning strategy for different types of tasks and scenarios in computational pathology.

Nowadays, gastric cancer has become a significant issue in the global cancer burden, and its impact cannot be ignored. The rapid development of artificial intelligence technology is attempting to address this situation, aiming to change the clinical management landscape of gastric cancer fundamentally. In this transformative change, machine learning and deep learning, as two core technologies, play a pivotal role, bringing unprecedented innovations and breakthroughs in the diagnosis, treatment, and prognosis evaluation of gastric cancer. This article comprehensively reviews the latest research status and application of artificial intelligence algorithms in gastric cancer, covering multiple dimensions such as image recognition, pathological analysis, personalized treatment, and prognosis risk assessment. These applications not only significantly improve the sensitivity of gastric cancer risk monitoring, the accuracy of diagnosis, and the precision of survival prognosis but also provide robust data support and a scientific basis for clinical decision-making. The integration of artificial intelligence, from optimizing the diagnosis process and enhancing diagnostic efficiency to promoting the practice of precision medicine, demonstrates its promising prospects for reshaping the treatment model of gastric cancer. Although most of the current AI-based models have not been widely used in clinical practice, with the continuous deepening and expansion of precision medicine, we have reason to believe that a new era of AI-driven gastric cancer care is approaching.

Deep learning has been successfully applied to histopathology image classification tasks. However, the performance of deep models is data-driven, and the acquisition and annotation of pathological image samples are difficult, which limit the model's performance. Compared to whole slide images (WSI) of patients, histopathology image datasets of animal models are easier to acquire and annotate. Therefore, this paper proposes an unsupervised domain adaptation method based on semantic correlation clustering for histopathology image classification. The aim is to utilize Minmice model histopathology image dataset to achieve the classification and recognition of human WSIs. Firstly, the multi-scale fused features extracted from the source and target domains are normalized and mapped. In the new feature space, the cosine distance between class centers is used to measure the semantic correlation between categories. Then, the domain centers, class centers, and sample distributions are self-constrainedly aligned. Multi-granular information is applied to achieve cross-domain semantic correlation knowledge transfer between classes. Finally, the probabilistic heatmap is used to visualize the model's prediction results and annotate the cancerous regions in WSIs. Experimental results show that the proposed method has high classification accuracy for WSI, and the annotated result is close to manual annotation, indicating its potential for clinical applications.

Digital pathology and artificial intelligence (AI) hold immense transformative potential to revolutionize cancer diagnostics, treatment outcomes, and biomarker discovery. Gaining a deeper understanding of deep learning algorithm methods applied to histopathological data and evaluating their performance on different tasks is crucial for developing the next generation of AI technologies. To this end, we developed AI in Histopathology Explorer (HistoPathExplorer); an interactive dashboard with intelligent tools available at www.histopathexpo.ai . This real-time online resource enables users, including researchers, decision-makers, and various stakeholders, to assess the current landscape of AI applications for specific clinical tasks, analyze their performance, and explore the factors influencing their translation into practice. Moreover, a quality index was defined for evaluating the comprehensiveness of methodological details in published AI methods. HistoPathExplorer highlights opportunities and challenges for AI in histopathology, and offers a valuable resource for creating more effective methods and shaping strategies and guidelines for translating digital pathology applications into clinical practice.

Methylation profiling of central nervous system (CNS) tumors, pioneered by the German Cancer Research Center, has significantly improved diagnostic accuracy. This study aimed to further enhance the performance of methylation classifiers by leveraging publicly available data and innovative machine-learning techniques.

Seoul National University Hospital Methylation Classifier (SNUH-MC) addressed data imbalance using the Synthetic Minority Over-sampling Technique (SMOTE) algorithm and incorporated OpenMax within a Multi-Layer Perceptron to prevent labeling errors in low-confidence diagnoses. Compared to two published CNS tumor methylation classification models (DKFZ-MC: Deutsches Krebsforschungszentrum Methylation Classifier v11b4: RandomForest, 767-MC: Multi-Layer Perceptron), our SNUH-MC showed improved performance in F1-score. For 'Filtered Test Data Set 1,' the SNUH-MC achieved higher F1-micro (0.932) and F1-macro (0.919) scores compared to DKFZ-MC v11b4 (F1-micro: 0.907, F1-macro: 0.627). We evaluated the performance of three classifiers; SNUH-MC, DKFZ-MC v11b4, and DKFZ-MC v12.5, using specific criteria. We set established 'Decisions' categories based on histopathology, clinical information, and next-generation sequencing to assess the classification results. When applied to 193 unknown SNUH methylation data samples, SNUH-MC notably improved diagnosis compared to DKFZ-MC v11b4. Specifically, 17 cases were reclassified as 'Match' and 34 cases as 'Likely Match' when transitioning from DKFZ-MC v11b4 to SNUH-MC. Additionally, SNUH-MC demonstrated similar results to DKFZ-MC v12.5 for 23 cases that were unclassified by v11b4.

This study presents SNUH-MC, an innovative methylation-based classification tool that significantly advances the field of neuropathology and bioinformatics. Our classifier incorporates cutting-edge techniques such as the SMOTE and OpenMax resulting in improved diagnostic accuracy and robustness, particularly when dealing with unknown or noisy data.

Colorectal cancer (CRC) is the third most common and second most deadly cancer worldwide. Despite advances in screening and treatment, CRC is heterogeneous and the response to therapy varies significantly, limiting personalized treatment options. Certain molecular biomarkers, including microsatellite instability (MSI), are critical in planning personalized treatment, although only a subset of patients may benefit. Currently, the primary methods for assessing MSI status include immunohistochemistry (IHC) for DNA mismatch repair proteins (MMRs), polymerase chain reaction (PCR)-based molecular testing, or next-generation sequencing (NGS). However, these techniques have limitations, are expensive and time-consuming, and often result in inter-method inconsistencies. Deficient mismatch repair (dMMR) or high microsatellite instability (MSI-H) are critical predictive biomarkers of response to immune checkpoint inhibitor (ICI) therapy and MSI testing is recommended to identify patients who may benefit. There is a pressing need for a more robust, reliable, and cost-effective approach that accurately assesses MSI status. Recent advances in computational pathology, in particular the development of technologies that digitally scan whole slide images (WSI) at high resolution, as well as new approaches to artificial intelligence (AI) in medicine, are increasingly gaining ground. This review aims to provide an overview of the latest findings on WSI and advances in AI methods for predicting MSI status, summarize their applications in CRC, and discuss their strengths and limitations in daily clinical practice.

Digital Pathology (DP) revolutionizes the diagnostic workflow. Digitized scanned slides enhance operational efficiency by facilitating remote access, slide storage, reporting and automated AI image analysis, and enabling collaboration and research. However, substantial upfront and maintenance costs remain significant barriers to adoption. This study evaluates DP's financial and qualitative value, exploring whether the long-term financial benefits justify investments and addressing implementation challenges in large public and private European laboratory settings. A targeted literature review, semi-structured interviews, surveys, and a net present value (NPV) model were employed to assess DP's impact on clinical practice and laboratory financials. Qualitative findings validate the key benefits of DP, including optimized workflow, enhanced logistics, and improved laboratory organization. Pathologists reported a smooth integration, improved training, teaching, and research capabilities, and increased flexibility through remote work. Collaboration within multidisciplinary teams was strengthened, while case examination efficiency and access to archival slides were notably improved. Quantitative results indicate that DP demonstrates strong financial potential, achieving cost recovery within 6 years. DP investment results in a 7-year NPV of + €0.21 million (m) driven by increased productivity and diagnosis volumes. Although the high upfront costs for scanners, training, and system integration pose a significant barrier to the adoption of DP, larger institutions are better positioned to leverage economies of scale. This study underscores the importance of sustained financial support to cope with the initial investment and regional collaboration in driving widespread adoption of DP. Expanding reimbursement policies for pathology procedures could significantly reduce financial barriers.

Artificial intelligence (AI) holds the potential to transform the management of upper gastrointestinal (GI) conditions, such as Barrett's esophagus, esophageal squamous cell cancer, and early gastric cancer. Advancements in deep learning and convolutional neural networks offer improved diagnostic accuracy and reduced diagnostic variability across different clinical settings, particularly where human error or fatigue may impair diagnostic precision. Deep learning models have shown the potential to improve early cancer detection and lesion characterization, predict invasion depth, and delineate lesion margins with remarkable accuracy, all contributing to effective treatment planning. Several challenges, however, limit the broad application of AI in GI endoscopy, particularly in the upper GI tract. Subtle lesion morphology and restricted diversity in training datasets, which are often sourced from specialized centers, may constrain the generalizability of AI models in various clinical settings. Furthermore, the "black box" nature of some AI systems can impede explainability and clinician trust. To address these issues, efforts are underway to incorporate multimodal data, such as combining endoscopic and histopathologic imaging, to bolster model robustness and transparency. In the future, AI promises substantial advancements in automated real-time endoscopic guidance, personalized risk assessment, and optimized biopsy decision making. As it evolves, it would substantially impact not only early diagnosis and prognosis, but also the cost-effectiveness of managing upper GI diseases, ultimately leading to improved patient outcomes and more efficient health care delivery.

In less than a decade, immune checkpoint inhibitors (ICIs) have transformed the management of mismatch repair-deficient (dMMR) and microsatellite instability-high (MSI) cancers. However, beyond colorectal cancer (CRC), much of the evidence is mostly derived from non-randomized phase II studies or post-hoc analyses of broader clinical trials. dMMR/MSI tumours represent a specific subgroup of gastro-esophageal adenocarcinomas (GEA), accounting for approximately 9 % of cases, with a higher prevalence in early-stage compared to advanced-stage disease and older female patients. These tumours are predominantly sporadic, often linked to MLH1 promoter methylation, and rarely exhibit HER2 overexpression/ERBB2 amplification or other oncogenic drivers. The treatment landscape for early stage dMMR/MSI GEA is likely to change substantially soon, as ICIs have shown high pathological complete response (pCR) rates in small phase II trials, raising questions on optimisation of neoadjuvant therapy, and paving the way for organ preservation. The standard of treatment for untreated patients with advanced dMMR/MSI GEA is chemotherapy + ICI irrespectively of PDL-1 status. However, the role of chemotherapy-free regimen consisting of CTLA-4 plus PD-1 inhibitors remains undetermined. This review addresses these and other emerging questions, offering expert opinions and insights into the future therapeutic landscape for dMMR/MSI GEA.

The development of new therapeutic strategies such as immune checkpoint inhibitors (ICIs) and targeted therapies has increased the complexity of the treatment landscape for solid tumors. At the current rate of annual FDA approvals, the potential treatment options could increase by tenfold over the next 5 years. The cost of personalized medicine technologies limits its accessibility, thus increasing socioeconomic disparities in the treated population. In this review we describe artificial intelligence (AI)-based solutions - including deep learning (DL) methods for routine medical imaging and large language models (LLMs) for electronic health records (EHRs) - to support cancer treatment decisions with cost-effective biomarkers. We address the current limitations of these technologies and propose the next steps towards their adoption in routine clinical practice.

Human epidermal growth factor receptor 2 (HER2) positive gastric cancer (GC) shows a robust response to the combined therapy based HER2-targeted therapy. The application of these therapies is highly dependent on the evaluation of tumor HER2 status. However, there are many risks and challenges in HER2 assessment in GC. Therefore, an economically viable and readily available instrument is requisite for distinguishing HER2 status among patients diagnosed with GC. The study has innovatively developed a deep learning model, HER2Net, which can predict the HER2 status by quantitatively calculating the proportion of HER2 high-expression regions. The HER2Net is trained on an internal training set derived from 531 hematoxylin & eosin (H&E) whole slide images (WSI) of 520 patients. Subsequently, the performance of HER2Net is validated on an internal test set from 115 H&E WSI of 111 patients and an external multi-center test set from 102 H&E WSI of 101 patients. The HER2Net achieves an accuracy of 0.9043 on the internal test set, and an accuracy of 0.8922 on an external test set from multiple institutes. This discovery indicates that the HER2Net can potentially offer a novel methodology for the identification of HER2-positive GC.

Training machine-learning models with synthetically generated data can alleviate the problem of data scarcity when acquiring diverse and sufficiently large datasets is costly and challenging. Here we show that cascaded diffusion models can be used to synthesize realistic whole-slide image tiles from latent representations of RNA-sequencing data from human tumours. Alterations in gene expression affected the composition of cell types in the generated synthetic image tiles, which accurately preserved the distribution of cell types and maintained the cell fraction observed in bulk RNA-sequencing data, as we show for lung adenocarcinoma, kidney renal papillary cell carcinoma, cervical squamous cell carcinoma, colon adenocarcinoma and glioblastoma. Machine-learning models pretrained with the generated synthetic data performed better than models trained from scratch. Synthetic data may accelerate the development of machine-learning models in scarce-data settings and allow for the imputation of missing data modalities.

CRC remains a significant public health challenge due to its high prevalence and mortality rates. Emerging evidence highlights the critical role of the gut microbiota in both the pathogenesis of CRC and the efficacy of treatment strategies, including chemotherapy and immunotherapy. Dysbiosis, characterized by imbalances in microbial communities, has been implicated in CRC progression and therapeutic outcomes. This review examines the intricate relationship between gut microbiota composition and CRC, emphasizing the potential for microbial profiles to serve as biomarkers for early detection and prognosis. Various interventions, such as prebiotics, probiotics, postbiotics, fecal microbiota transplantation, and dietary modifications, aim to restore microbiota balance and shift dysbiosis toward eubiosis, thereby improving health outcomes. Additionally, the integration of microbial profiling into clinical practice could enhance diagnostic capabilities and personalize treatment strategies, advancing the field of oncology. The study of intratumoral microbiota offers new diagnostic and prognostic tools that, combined with artificial intelligence algorithms, could predict treatment responses and assess the risk of adverse effects. Given the growing understanding of the gut microbiome-cancer axis, developing microbiota-oriented strategies for CRC prevention and treatment holds promise for improving patient care and clinical outcomes.

Inferring cell spatial profiles from histology images is critical for cancer diagnosis and treatment in clinical settings. In this study, we report a weakly-supervised deep-learning method, HistoCell, to directly infer super-resolution cell spatial profiles consisting of cell types, cell states and their spatial network from histology images at the single-nucleus-level. Benchmark analysis demonstrates that HistoCell robustly achieves state-of-the-art performance in terms of cell type/states prediction solely from histology images across multiple cancer tissues. HistoCell can significantly enhance the deconvolution accuracy for the spatial transcriptomics data and enable accurate annotation of subtle cancer tissue architectures. Moreover, HistoCell is applied to de novo discovery of clinically relevant spatial organization indicators, including prognosis and drug response biomarkers, across diverse cancer types. HistoCell also enable image-based screening of cell populations that drives phenotype of interest, and is applied to discover the cell population and corresponding spatial organization indicators associated with gastric malignant transformation risk. Overall, HistoCell emerges as a powerful and versatile tool for cancer studies in histology image-only cohorts.

Accumulated evidence has implicated the diverse and substantial influence of lactate on cellular differentiation and fate regulation in physiological and pathological settings, particularly in intricate conditions such as cancer. Specifically, lactate has been demonstrated to be pivotal in molding the tumor microenvironment (TME) through its effects on different cell populations. Within tumor cells, lactate impacts cell signaling pathways, augments the lactate shuttle process, boosts resistance to oxidative stress, and contributes to lactylation. In various cellular populations, the interplay between lactate and immune cells governs processes such as cell differentiation, immune response, immune surveillance, and treatment effectiveness. Furthermore, communication between lactate and stromal/endothelial cells supports basal membrane (BM) remodeling, epithelial-mesenchymal transitions (EMT), metabolic reprogramming, angiogenesis, and drug resistance. Focusing on lactate production and transport, specifically through lactate dehydrogenase (LDH) and monocarboxylate transporters (MCT), has shown promise in the treatment of cancer. Inhibitors targeting LDH and MCT act as both tumor suppressors and enhancers of immunotherapy, leading to a synergistic therapeutic effect when combined with immunotherapy. The review underscores the importance of lactate in tumor progression and provides valuable perspectives on potential therapeutic approaches that target the vulnerability of lactate metabolism, highlighting the Heel of Achilles for cancer treatment.

Deep Learning (DL) has emerged as a powerful tool to predict genetic biomarkers directly from digitized Hematoxylin and Eosin (H&E) slides in colorectal cancer (CRC). However, few studies have systematically investigated the predictability of biomarkers beyond routinely available alterations such as microsatellite instability (MSI), and BRAF and KRAS mutations.

Our primary dataset comprised H&E slides of CRC tumors across five cohorts totaling 1,376 patients who underwent comprehensive panel sequencing, with an additional 536 patients from two public datasets for validation. We developed a DL model using a single transformer model to predict multiple genetic alterations directly from the slides. The model's performance was compared against conventional single-target models, and potential confounders were analyzed.

The multi-target model was able to predict numerous biomarkers from pathology slides, matching and partly exceeding single-target transformers. The Area Under the Receiver Operating Characteristic curve (AUROC, mean ± std) on the primary external validation cohorts was: BRAF (0·78 ± 0·01), hypermutation (0·88 ± 0·01), MSI (0·93 ± 0·01), RNF43 (0·86 ± 0·01); this biomarker predictability was mirrored across metrics and co-occurrence analyses. However, biomarkers with high AUROCs largely correlated with MSI, with model predictions depending considerably on MSI-associated morphology upon pathological examination.

Our study demonstrates that multi-target transformers can predict the biomarker status for numerous genetic alterations in CRC directly from H&E slides. However, their predictability is mainly associated with MSI phenotype, despite indications of slight biomarker-inherent contributions to a phenotype. Our findings underscore the need to analyze confounders in AI-based oncology biomarkers. To enable this, we developed a validated model applicable to other cancers and larger, diverse datasets.

The German Federal Ministry of Health, the Max-Eder-Programme of German Cancer Aid, the German Federal Ministry of Education and Research, the German Academic Exchange Service, and the EU.

Accurate segmentation and classification of cell nuclei are crucial for histopathological image analysis. However, existing deep neural network-based methods often struggle to capture complex morphological features and global spatial distributions of cell nuclei due to their reliance on local receptive fields.

This study proposes a graph neural structure encoding framework based on a vision-language model. The framework incorporates: (1) A multi-scale feature fusion and knowledge distillation module utilizing the Contrastive Language-Image Pre-training (CLIP) model's image encoder; (2) A method to transform morphological features of cells into textual descriptions for semantic representation; and (3) A graph neural network approach to learn spatial relationships and contextual information between cell nuclei.

Experimental results demonstrate that the proposed method significantly improves the accuracy of cell nucleus segmentation and classification compared to existing approaches. The framework effectively captures complex nuclear structures and global distribution features, leading to enhanced performance in histopathological image analysis.

By deeply mining the morphological features of cell nuclei and their spatial topological relationships, our graph neural structure encoding framework achieves high-precision nuclear segmentation and classification. This approach shows significant potential for enhancing histopathological image analysis, potentially leading to more accurate diagnoses and improved understanding of cellular structures in pathological tissues.

Endometrial cancer is a prevalent gynecological malignancy that requires accurate identification of its molecular subtypes for effective diagnosis and treatment. Four molecular subtypes with different clinical outcomes have been identified: POLE mutation, mismatch repair deficient, p53 abnormal, and no specific molecular profile. However, determining these subtypes typically relies on expensive gene sequencing. To overcome this limitation, we propose a novel method that utilizes hematoxylin and eosin-stained whole slide images to predict endometrial cancer molecular subtypes.

Our approach leverages a hierarchical foundation model as a backbone, fine-tuned from the UNI computational pathology foundation model, to extract tissue embedding from different scales. We have achieved promising results through extensive experimentation on the Fudan University Shanghai Cancer Center cohort (N = 364). Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853-0.904) in a five-fold cross-validation. Compared to the current state-of-the-art molecular subtypes prediction for endometrial cancer, our method outperforms in terms of predictive accuracy and computational efficiency. Moreover, our method is highly reproducible, allowing for ease of implementation and widespread adoption. This study aims to address the cost and time constraints associated with traditional gene sequencing techniques. By providing a reliable and accessible alternative to gene sequencing, our method has the potential to revolutionize the field of endometrial cancer diagnosis and improve patient outcomes.

The codes and data used for generating results in this study are available at https://github.com/HaoyuCui/hi-UNI for GitHub and https://doi.org/10.5281/zenodo.14627478 for Zenodo.

Biliary abnormalities in autoimmune hepatitis (AIH) and interface hepatitis in primary biliary cholangitis (PBC) occur frequently, and misinterpretation may lead to therapeutic mistakes with a negative impact on patients. This study investigates the use of a deep learning (DL)-based pipeline for the diagnosis of AIH and PBC to aid differential diagnosis.

We conducted a multicenter study across six European referral centers, and built a library of digitized liver biopsy slides dating from 1997 to 2023. A training set of 354 cases (266 AIH and 102 PBC) and an external validation set of 92 cases (62 AIH and 30 PBC) were available for analysis. A novel DL model, the autoimmune liver neural estimator (ALNE), was trained on whole-slide images (WSIs) with H&E staining, without human annotations. The ALNE model was evaluated against clinico-pathological diagnoses and tested for interobserver variability among general pathologists.

The ALNE model demonstrated high accuracy in differentiating AIH from PBC, achieving an area under the receiver operating characteristic curve of 0.81 in external validation. Attention heatmaps showed that ALNE tends to focus more on areas with increased inflammation, associating such patterns predominantly with AIH. A multivariate explainable ML model revealed that PBC cases misclassified as AIH more often had ALP values between 1 × upper limit of normal (ULN) and 2 × ULN, coupled with AST values above 1 × ULN. Inconsistency among general pathologists was noticed when evaluating a random sample of the same cases (Fleiss's kappa value 0.09).

The ALNE model is the first system generating a quantitative and accurate differential diagnosis between cases with AIH or PBC.

This study demonstrates the significant potential of the autoimmune liver neural estimator model, a transformer-based deep learning system, in accurately distinguishing between autoimmune hepatitis and primary biliary cholangitis using digitized liver biopsy slides without human annotation. The scientific justification for this work lies in addressing the challenge of differentiating these conditions, which often present with overlapping features and can lead to therapeutic mistakes. In addition, there is need for quantitative assessment of information embedded in liver biopsies, which are currently evaluated on qualitative or semi-quantitative methods. The results of this study are crucial for pathologists, researchers, and clinicians, providing a reliable diagnostic tool that reduces interobserver variability and improves diagnostic accuracy of these conditions. Potential methodological limitations, such as the diversity in scanning techniques and slide colorations, were considered, ensuring the robustness and generalizability of the findings.

Treatment approaches for colorectal cancer (CRC) are highly dependent on the molecular subtype, as immunotherapy has shown efficacy in cases with microsatellite instability (MSI) but is ineffective for the microsatellite stable (MSS) subtype. There is promising potential in utilizing deep neural networks (DNNs) to automate the differentiation of CRC subtypes by analyzing hematoxylin and eosin (H&E) stained whole-slide images (WSIs). Due to the extensive size of WSIs, multiple instance learning (MIL) techniques are typically explored. However, existing MIL methods focus on identifying the most representative image patches for classification, which may result in the loss of critical information. Additionally, these methods often overlook clinically relevant information, like the tendency for MSI class tumors to predominantly occur on the proximal (right side) colon.

We introduce 'CIMIL-CRC', a DNN framework that: (1) solves the MSI/MSS MIL problem by efficiently combining a pre-trained feature extraction model with principal component analysis (PCA) to aggregate information from all patches, and (2) integrates clinical priors, particularly the tumor location within the colon, into the model to enhance patient-level classification accuracy. We assessed our CIMIL-CRC method using the average area under the receiver operating characteristic curve (AUROC) from a 5-fold cross-validation experimental setup for model development on the TCGA-CRC-DX cohort, contrasting it with a baseline patch-level classification, a MIL-only approach, and a clinically-informed patch-level classification approach.

Our CIMIL-CRC outperformed all methods (AUROC: 0.92±0.002 (95% CI 0.91-0.92), vs. 0.79±0.02 (95% CI 0.76-0.82), 0.86±0.01 (95% CI 0.85-0.88), and 0.87±0.01 (95% CI 0.86-0.88), respectively). The improvement was statistically significant. To the best of our knowledge, this is the best result achieved for MSI/MSS classification on this dataset.

Our CIMIL-CRC method holds promise for offering insights into the key representations of histopathological images and suggests a straightforward implementation.

Deficient mismatch repair or microsatellite instability is a major predictive biomarker for the efficacy of immune checkpoint inhibitors of colorectal cancer. However, routine testing has not been uniformly implemented due to cost and resource constraints.

We developed and validated a deep learning-based classifiers to detect mismatch repair-deficient status from routine colonoscopy images. We obtained the colonoscopy images from the imaging database at Endoscopic Center of the Sixth Affiliated Hospital, Sun Yat-sen University. Colonoscopy images from a prospective trial (Neoadjuvant PD-1 blockade by toripalimab with or without celecoxib in mismatch repair-deficient or microsatellite instability-high locally advanced colorectal cancer) were used to test the model.

A total of 5226 eligible images from 892 tumors from the consecutive patients were utilized to develop and validate the deep learning model. 2105 colorectal cancer images from 306 tumors were randomly selected to form model development dataset with a class-balanced approach. 3121 images of 488 proficient mismatch repair tumors and 98 deficient mismatch repair tumors were used to form the independent dataset. The model achieved an AUROC of 0.948 (95% CI 0.919-0.977) on the test dataset. On the independent validation dataset, the AUROC was 0.807 (0.760-0.854), and the NPV in was 94.2% (95% CI 0.918-0.967). On the prospective trial dataset, the model identified 29 tumors among the 33 deficient mismatch repair tumors (87.88%).

The model achieved a high NPV in detecting deficient mismatch repair colorectal cancers. This model might serve as an automatic screening tool.

In clinical practice, frozen section (FS) images can be utilized to obtain the immediate pathological results of the patients in operation due to their fast production speed. However, compared with the formalin-fixed and paraffin-embedded (FFPE) images, the FS images greatly suffer from poor quality. Thus, it is of great significance to transfer the FS image to the FFPE one, which enables pathologists to observe high-quality images in operation. However, obtaining the paired FS and FFPE images is quite hard, so it is difficult to obtain accurate results using supervised methods. Apart from this, the FS to FFPE stain transfer faces many challenges. Firstly, the number and position of nuclei scattered throughout the image are hard to maintain during the transfer process. Secondly, transferring the blurry FS images to the clear FFPE ones is quite challenging. Thirdly, compared with the center regions of each patch, the edge regions are harder to transfer. To overcome these problems, a multi-perspective self-supervised GAN, incorporating three auxiliary tasks, is proposed to improve the performance of FS to FFPE stain transfer. Concretely, a nucleus consistency constraint is designed to enable the high-fidelity of nuclei, an FFPE guided image deblurring is proposed for improving the clarity, and a multi-field-of-view consistency constraint is designed to better generate the edge regions. Objective indicators and pathologists' evaluation for experiments on the five datasets across different countries have demonstrated the effectiveness of our method. In addition, the validation in the downstream task of microsatellite instability prediction has also proved the performance improvement by transferring the FS images to FFPE ones. Our code link is https://github.com/linyiyang98/Self-Supervised-FS2FFPE.git.

Targeted therapy for intrahepatic cholangiocarcinoma (ICC) shows superior survival outcomes but patients with certain targetable alterations are no more than 20%. Genetic alteration screening for all ICC patients is of high cost and not routinely performed. This study intends to develop a histopathology-based artificial intelligence (AI)-assisted system for predicting genetic alteration of ICC.

We constructed a Genetic Alteration Prediction (GAP) system based on multi-instance learning and self-supervised learning to predict genetic alterations using whole-slide images (WSIs) of H&E-stained slides. A total of 2069 WSIs from 232 ICC patients underwent surgery of the FAH-SYSU dataset were used for model construction and adjustment by five-fold cross-validation. Another 150 patients from three medical centres were used as independent external validations. We also compared the cost-effectiveness of GAP-assisted precise treatment and all-sequencing strategy to non-sequencing strategy.

The GAP was able to predict actionable genetic alterations of ICC, including FGFR2 and IDH. The area under the receiver operating characteristic curves (AUC) for FGFR2 and IDH were 0.754 and 0.713 in the internal dataset, and 0.724 and 0.656 in the external dataset, respectively. Furthermore, compared to giving chemotherapy without sequencing for every patient, GAP-assisted precise treatment could increase 1 progression-free quality-adjusted life month with a cost of $13871.72, the co-responding figure for all-sequencing strategy is $44538.93. Decision curve analysis showed that AI-assisted strategy provides better clinical benefits.

We constructed an AI-assisted genetic alteration screening system which is predictable to ICC actionable targets and has potential to assist precise targeted treatment of advanced ICC.

Only a small fraction of patients with advanced non-small cell lung cancer (NSCLC) respond to immune checkpoint inhibitor (ICI) treatment. For optimal personalized NSCLC care, it is imperative to identify patients who are most likely to benefit from immunotherapy.

To develop a supervised deep learning-based ICI response prediction method; evaluate its performance alongside other known predictive biomarkers; and assess its association with clinical outcomes in patients with advanced NSCLC.

This multicenter cohort study developed and independently validated a deep learning-based response stratification model for predicting ICI treatment outcome in patients with advanced NSCLC from whole slide hematoxylin and eosin-stained images. Images for model development and validation were obtained from 1 participating center in the US and 3 in the European Union (EU) from August 2014 to December 2022. Data analyses were performed from September 2022 to May 2024.

Monotherapy with ICIs.

Model performance measured by clinical end points and objective response rate (ORR) differentiation power vs other predictive biomarkers, ie, programmed death-ligand 1 (PD-L1), tumor mutational burden (TMB), and tumor-infiltrating lymphocytes (TILs).

A total of 295 581 image tiles from 958 patients (mean [SD] age, 66.0 [10.6] years; 456 [48%] females and 502 [52%] males) treated with ICI for NSCLC were included in the analysis. The US-based development cohort consisted of 614 patients with median (IQR) follow-up time of 54.5 (38.2-68.1) months, and the EU-based validation cohort, 344 patients with 43.3 (27.4-53.9) months of follow-up. The ORR to ICI was 26% in the developmental cohort and 28% in the validation cohort. The deep learning model's area under the receiver operating characteristic curve (AUC) for ORR was 0.75 (95% CI, 0.64-0.85) in the internal test set and 0.66 (95% CI, 0.60-0.72) in the validation cohort. In a multivariable analysis, the deep learning model's score was an independent predictor of ICI response in the validation cohort for both progression-free (hazard ratio, 0.56; 95% CI, 0.42-0.76; P < .001) and overall survival (hazard ratio, 0.53; 95% CI, 0.39-0.73; P < .001). The tuned deep learning model achieved a higher AUC than TMB, TILs, and PD-L1 in the internal set; in the validation cohort, it was superior to TILs and comparable with PD-L1 (AUC, 0.67; 95% CI, 0.60-0.74), with a 10-percentage point improvement in specificity. In the validation cohort, combining the deep learning model with PD-L1 scores achieved an AUC of 0.70 (95% CI, 0.63-0.76), outperforming either marker alone, with a response rate of 51% compared to 41% for PD-L1 (≥50%) alone.

The findings of this cohort study demonstrate a strong and independent deep learning-based feature associated with ICI response in patients with NSCLC across various cohorts. Clinical use of this deep learning model could refine treatment precision and better identify patients who are likely to benefit from ICI for treatment of advanced NSCLC.

The confluence of new technologies with artificial intelligence (AI) and machine learning (ML) analytical techniques is rapidly advancing the field of precision oncology, promising to improve diagnostic approaches and therapeutic strategies for patients with cancer. By analyzing multi-dimensional, multiomic, spatial pathology, and radiomic data, these technologies enable a deeper understanding of the intricate molecular pathways, aiding in the identification of critical nodes within the tumor's biology to optimize treatment selection. The applications of AI/ML in precision oncology are extensive and include the generation of synthetic data, e.g., digital twins, in order to provide the necessary information to design or expedite the conduct of clinical trials. Currently, many operational and technical challenges exist related to data technology, engineering, and storage; algorithm development and structures; quality and quantity of the data and the analytical pipeline; data sharing and generalizability; and the incorporation of these technologies into the current clinical workflow and reimbursement models.

This study aimed to develop a multi-modality model by incorporating pretreatment computed tomography (CT) radiomics and pathomics features along with clinical variables to predict pathologic complete response (pCR) to neoadjuvant chemoimmunotherapy in patients with locally advanced esophageal cancer (EC).

A total of 223 EC patients who underwent neoadjuvant chemoimmunotherapy followed by surgical intervention between August 2021 and December 2023 were included in this study. Radiomics features were extracted from contrast-enhanced CT images using PyrRadiomics, while pathomics features were derived from whole-slide images (WSIs) of pathological specimens using a fine-tuned deep learning model (ResNet-50). After feature selection, three single-modality prediction models and a combined multi-modality model integrating two radiomics features, 11 pathomics features, and two clinicopathological features were constructed using the support vector machine (SVM) algorithm. The performance of the models were evaluated using receiver operating characteristic (ROC) analysis, calibration plots, and decision curve analysis (DCA). Shapley values were also utilized to explain the prediction model.

The predictive capability of the multi-modality model in predicting pCR yielded an area under the curve (AUC) of 0.89 (95% confidence interval [CI], 0.75-1.00), outperforming the radiomics model (AUC 0.70 [95% CI 0.54-0.85]), pathomics model (AUC 0.77 [95% CI 0.53-1.00]), and clinical model (AUC 0.63 [95% CI 0.46-0.80]). Additionally, both the calibration plot and DCA curves support the clinical utility of the integrated multi-modality model.

The combined multi-modality model we propose can better predict the pCR status of esophageal cancer and help inform clinical treatment decisions.

Chat Generative Pretrained Transformer (ChatGPT) is a type of large language model (LLM) developed by OpenAI, known for its extensive knowledge base and interactive capabilities. These attributes make it a valuable tool in the medical field, particularly for tasks such as answering medical questions, drafting clinical notes, and optimizing the generation of radiology reports. However, keeping accuracy in medical contexts is the biggest challenge to employing GPT-4 in a clinical setting. This study aims to investigate the accuracy of GPT-4, which can process both text and image inputs, in generating diagnoses from pathological images.

This study analyzed 44 histopathological images from 16 organs and 100 colorectal biopsy photomicrographs. The initial evaluation was conducted using the standard GPT-4 model in January 2024, with a subsequent re-evaluation performed in July 2024. The diagnostic accuracy of GPT-4 was assessed by comparing its outputs to a reference standard using statistical measures. Additionally, four pathologists independently reviewed the same images to compare their diagnoses with the model's outputs. Both scanned and photographed images were tested to evaluate GPT-4's generalization ability across different image types.

GPT-4 achieved an overall accuracy of 0.64 in identifying tumor imaging and tissue origins. For colon polyp classification, accuracy varied from 0.57 to 0.75 in different subtypes. The model achieved 0.88 accuracy in distinguishing low-grade from high-grade dysplasia and 0.75 in distinguishing high-grade dysplasia from adenocarcinoma, with a high sensitivity in detecting adenocarcinoma. Consistency between initial and follow-up evaluations showed slight to moderate agreement, with Kappa values ranging from 0.204 to 0.375.

GPT-4 demonstrates the ability to diagnose pathological images, showing improved performance over earlier versions. Its diagnostic accuracy in cancer is comparable to that of pathology residents. These findings suggest that GPT-4 holds promise as a supportive tool in pathology diagnostics, offering the potential to assist pathologists in routine diagnostic workflows.

In recent years, novel findings have progressively and promisingly supported the potential role of Artificial intelligence (AI) in transforming the management of various neoplasms, including hepatocellular carcinoma (HCC). HCC represents the most common primary liver cancer. Alarmingly, the HCC incidence is dramatically increasing worldwide due to the simultaneous "pandemic" spreading of metabolic dysfunction-associated steatotic liver disease (MASLD). MASLD currently constitutes the leading cause of chronic hepatic damage (steatosis and steatohepatitis), fibrosis, and liver cirrhosis, configuring a scenario where an HCC onset has been reported even in the early disease stage. On the other hand, HCC represents a serious plague, significantly burdening the outcomes of chronic hepatitis B (HBV) and hepatitis C (HCV) virus-infected patients. Despite the recent progress in the management of this cancer, the overall prognosis for advanced-stage HCC patients continues to be poor, suggesting the absolute need to develop personalized healthcare strategies further. In this "cold war", machine learning techniques and neural networks are emerging as weapons, able to identify the patterns and biomarkers that would have normally escaped human observation. Using advanced algorithms, AI can analyze large volumes of clinical data and medical images (including routinely obtained ultrasound data) with an elevated accuracy, facilitating early diagnosis, improving the performance of predictive models, and supporting the multidisciplinary (oncologist, gastroenterologist, surgeon, radiologist) team in opting for the best "tailored" individual treatment. Additionally, AI can significantly contribute to enhancing the effectiveness of metabolomics-radiomics-based models, promoting the identification of specific HCC-pathogenetic molecules as new targets for realizing novel therapeutic regimens. In the era of precision medicine, integrating AI into routine clinical practice appears as a promising frontier, opening new avenues for liver cancer research and treatment.

Cancer remains a leading cause of morbidity and mortality worldwide. Traditional treatments like chemotherapy and radiation often result in significant side effects and varied patient outcomes. Immunotherapy has emerged as a promising alternative, harnessing the immune system to target cancer cells. However, the complexity of immune responses and tumor heterogeneity challenges its effectiveness.

This mini-narrative review explores the role of artificial intelligence [AI] in enhancing the efficacy of cancer immunotherapy, predicting patient responses, and discovering novel therapeutic targets.

A comprehensive review of the literature was conducted, focusing on studies published between 2010 and 2024 that examined the application of AI in cancer immunotherapy. Databases such as PubMed, Google Scholar, and Web of Science were utilized, and articles were selected based on relevance to the topic.

AI has significantly contributed to identifying biomarkers that predict immunotherapy efficacy by analyzing genomic, transcriptomic, and proteomic data. It also optimizes combination therapies by predicting the most effective treatment protocols. AI-driven predictive models help assess patient response to immunotherapy, guiding clinical decision-making and minimizing side effects. Additionally, AI facilitates the discovery of novel therapeutic targets, such as neoantigens, enabling the development of personalized immunotherapies.

AI holds immense potential in transforming cancer immunotherapy. However, challenges related to data privacy, algorithm transparency, and clinical integration must be addressed. Overcoming these hurdles will likely make AI a central component of future cancer immunotherapy, offering more personalized and effective treatments.

Over the last decade, Hippo signaling has emerged as a major tumor-suppressing pathway. Its dysregulation is associated with abnormal expression of YAP1 and TEAD-family genes. Recent works have highlighted the role of YAP1/TEAD activity in several cancers and its potential therapeutic implications. Therefore, identifying patients with a dysregulated Hippo pathway is key to enhancing treatment impact. Although recent studies have derived RNA-seq-based signatures, there remains a need for a reproducible and cost-effective method to measure the pathway activation. In recent years, deep learning applied to histology slides have emerged as an effective way to predict molecular information from a data modality available in clinical routine. Here, we trained models to predict YAP1/TEAD activity from H&E-stained histology slides in multiple cancers. The robustness of our approach was assessed in seven independent validation cohorts. Finally, we showed that histological markers of disease aggressiveness were associated with dysfunctional Hippo signaling.

Whole slide imaging (WSI) has transformed diagnostic medicine, particularly in the field of cancer diagnosis and treatment. The use of deep learning algorithms for predicting WSIs has opened up new avenues for advanced medical diagnostics. Additionally, stain normalization can reduce the color and intensity variations present in WSI from different hospitals. As a result, deep learning classification accuracy improves. However, WSI reading and color normalization are still largely performed by using CPUs, leading to sub-optimal performance. We proposed a High-Performance Heterogeneous Computing Toolkit for Tissue Image (HCTTI) that integrates multiple computer system-level optimizations and encompasses WSI reading, tile normalization, and tile saving. We explored the potential advantages and limitations of different WSI readers and color normalization techniques in WSI analysis and the performance of different tile serialization formats for saving tiles. We found that HCTTI is 7 × faster than OpenSlide for reading WSIs, GPU implementation of the Macenko normalization algorithm is 9 × faster than TIAToolbox implementation, and HDF5 is faster than png and Zarr for storing normalized images in both writing (13 × acceleration compared to png) and reading (2 × acceleration compared to png), Specifically, HDF5 provides superior performance in handling large, complex datasets due to its efficient chunking and compression capabilities, as well as its broad support for hierarchical data management, making it very suitable for workloads like deep learning training I/O pattern that involves randomly reading large amount of small files in each training epoch. We also achieved linear acceleration in our multi-node distributed GPU implementation. To our knowledge, HCTTI is the first comprehensive toolkit that comprises distributed WSI reading, normalization, and serialization. It is 13 × speedup compared to TIAToolbox implementation for normalizing a single WSI. Our findings could help pave the way for more effective and efficient deep learning-based approaches to WSI analysis, with the potential to transform medical diagnosis and treatment for a wide range of conditions. The source code of HCTTI is available at https://github.com/wangbo00129/HCTTI .

Hemorrhagic transformation (HT) is a complication of reperfusion therapy following acute ischemic stroke (AIS). We aimed to develop and validate a model for predicting HT and its subtypes with poor prognosis-parenchymal hemorrhage (PH), including PH-1 (hematoma within infarcted tissue, occupying < 30%) and PH-2 (hematoma occupying ≥ 30% of the infarcted tissue)-in AIS patients following intravenous thrombolysis (IVT) based on noncontrast computed tomography (NCCT) and clinical data.

In this six-center retrospective study, clinical and imaging data from 445 consecutive IVT-treated AIS patients were collected (01/2018-06/2023). The training cohort comprised 344 patients from five centers, and the test cohort included 101 patients from the sixth center. A clinical model was developed using eXtreme Gradient Boosting, an NCCT-based imaging model was created using deep learning, and an ensemble model integrated both models. Comparison with existing clinical scores (MSS, SEDAN, GRASPS) was performed using the DeLong test.

Of the 445 individuals, 202 (45.4%) had HT, 79 (17.8%) had hemorrhagic infarction, and 123 (27.6%) had PH. In the test cohort, the area under the receiver operating characteristic curve (AUROC) of the clinical, imaging, and ensemble model for HT prediction was 0.877, 0.920, and 0.937, respectively. The ensemble model for HT prediction outperformed MSS, SEDAN, and GRASPS scores (p ≤ 0.023). The ensemble model predicted PH and PH-2 with AUROC of 0.858 and 0.806, respectively.

Developing and validating an integrated model that can predict HT and its subtypes in AIS patients following IVT based on NCCT and clinical data is feasible.

The clinical, imaging, and ensemble models based on noncontrast CT and clinical data outperformed existing clinical scores in predicting hemorrhagic transformation of AIS and its subtypes with poor prognosis, facilitating personalized treatment decisions.

The models demonstrated the capability to predict hemorrhagic transformation of acute ischemic stroke quickly, accurately, and reliably. The proposed models outperformed existing clinical scores in predicting hemorrhagic transformation. The ensemble model provided risk assessment of parenchymal hemorrhage and parenchymal hemorrhage-2 outperforming existing clinical scores.