J. N. Kather

Multimodal histopathologic models stratify hormone receptor-positive early breast cancer

This paper develops a multimodal transformer model using H&E images and text reports to predict recurrence scores in hormone receptor-positive breast cancer, achieving high accuracy and generalizability.

Validating a low-cost, open-source, locally manufactured workstation and computational pipeline for automated histopathology evaluation using deep learning

The paper validates a low-cost, open-source workstation and computational pipeline for automated histopathology evaluation using deep learning, showing comparable performance to expensive institutional hardware.

Latent transcriptional programs reveal histology-encoded tumor features spanning tissue origins

This paper uses machine learning to identify latent transcriptional programs in squamous cell carcinomas from histology images, linking them to clinically relevant biological processes.

Self-supervised deep learning for pan-cancer mutation prediction from histopathology

This paper develops a self-supervised deep learning pipeline to predict genetic mutations from histopathology images across seven cancer types in two large datasets.

Deep learning can predict multi-omic biomarkers from routine pathology images: A systematic large-scale study

This study shows deep learning can predict multi-omic biomarkers from H&E-stained cancer slides, with 50% of models achieving AUC > 0.633 across 32 cancer types.

DeepMed: A unified, modular pipeline for end-to-end deep learning in computational pathology

DeepMed provides a modular, scalable AI pipeline for predicting molecular and histopathological features from histopathology images, with public benchmark data.

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

The paper surveys computational pathology experts to identify promising AI applications in digital pathology, finding treatment response prediction and genetic alteration detection as top priorities.

Improving Mathematical Models of Cancer throughGame-Theoretic Modelling: A Study in Non-SmallCell Lung Cancer

This study applies a game-theoretic model to 590 Non-Small Cell Lung Cancer patients, showing it better fits tumor dynamics than classical models and predicts treatment outcomes.

Classical Mathematical Models for Prediction of Response to Chemotherapy and Immunotherapy

This study evaluates classical mathematical models for predicting chemotherapy and immunotherapy responses using real-world patient tumor volume data, finding the Gompertz model provides the best balance between fit and complexity.

Nerve fibers in the Tumor Microenvironment are co-localized with Tertiary Lymphoid Structures

The paper analyzes nerve fiber density and tertiary lymphoid structures in pancreatic cancer, finding that high nerve fiber density and TLS count correlate with improved survival.

Deep learning detects virus presence in cancer histology

This paper uses deep learning to detect the presence of HPV and EBV in cancer histology images, achieving high accuracy in predicting virus-associated tumors.

Spatial structure governs the mode of tumour evolution

The paper shows that tumour architecture determines evolutionary patterns, identifying four onco-evolutionary modes based on spatial structure and cell dispersal.

Harnessing patient-specific response dynamics to optimize evolutionary therapies for metastatic clear cell renal cell carcinoma - Learning to adapt

This paper develops a dynamic tumor-immune model to predict patient response to treatment using NLR as a biomarker and simulates optimal therapies for virtual RCC patients.