AI in cancer detection

Applying Self-Supervised Tissue Patch Encoders to Enable Cellular Graph Encoding

This study shows that a pretrained tissue patch encoder can effectively encode cellular graphs, capturing cell-level features and spatial relationships in tissue patches.

Detection and Classification of Cervical Cancer Using Optimized Deep Learning Approach

This paper develops a deep learning model for cervical cancer classification using Pap smear images, achieving high accuracy and outperforming previous methods.

Closing the gap in the clinical adoption of computational pathology: a standardized, open-source framework to integrate deep-learning algorithms into the laboratory information system

This paper presents a framework for integrating deep-learning models into the laboratory information system using open-source tools and HL7 standards to support clinical adoption in digital pathology.

Deep learning-based interpretable prediction of recurrence of diffuse large B-cell lymphoma

This study develops a deep learning model to predict DLBCL recurrence using histological images, achieving an F1 score of 0.83 and identifying predictive nuclear features.

MARQO pipeline resolves multiparametric cellular and spatial organization in cancer tissue lesions

MARQO pipeline automates single-cell analysis of multiplex immunostained tissue slides, showing improved accuracy over manual methods in cancer studies.

Versatile Stain Transfer in Histopathology Using a Unified Diffusion Framework

This paper introduces DUST, a diffusion model for histopathology stain transfer, showing improved performance over GAN-based methods on a kidney histopathological dataset.

Spatial mapping of immunosuppressive cancer-associated fibroblast gene signatures in H&E-stained images using additive multiple instance learning

This paper introduces aMIL, a machine learning method that predicts CAF gene signatures from H&E images, revealing spatial patterns linked to immunosuppression in tumors.

Deep Learning Predicts Survival Across Squamous Tumor Entities From H&E Stains: Insights from Head and Neck, Esophagus, Lung and Cervical Cancer

This study develops a deep learning model to predict survival across multiple squamous tumor types using H&E stained slides, showing its predictive power beyond head and neck cancer.

Extended pre-training of histopathology foundation models uncovers co-existing breast cancer archetypes characterized by RNA splicing or TGF-β dysregulation

This paper extends pre-training of histopathology foundation models to specialize them for tumor tissue, revealing two coexisting breast cancer archetypes defined by RNA splicing and TGF-β dysregulation.

Delta Marches: Generative AI based image synthesis to decode disease-driving morphologic transformations.

Delta-Marches uses generative AI to simulate morphological changes between disease classes, identifying subcellular features that drive disease progression in histopathological images.

Quantification of collagen and associated features from H&E-stained whole slide pathology images across cancer types using a physics-based deep learning model

This paper introduces a deep learning model to predict collagen from H&E-stained images, enabling collagen quantification across cancer types and linking it to clinical outcomes.

A foundation model for generalizable cancer diagnosis and survival prediction from histopathological images

This paper introduces BEPH, a lightweight foundation model for cancer diagnosis and survival prediction from histopathological images, showing improved performance over existing models with limited data.

Self-Supervised Learning Reveals Clinically Relevant Histomorphological Patterns for Therapeutic Strategies in Colon Cancer

This paper uses self-supervised learning to identify histomorphological patterns in colon cancer WSIs, revealing clinically relevant clusters associated with survival and treatment responses.

Self-supervised learning enables unbiased patient characterization from multiplexed cancer tissue microscopy images

The paper develops a self-supervised deep learning framework to analyze multiplexed cancer tissue microscopy images, stratifying patients into prognostically distinct groups with different survival outcomes.

Weakly Supervised Vector Quantization for Whole Slide Image Classification

This paper introduces VQ-MIL, a weakly supervised vector quantization method for WSI classification, achieving state-of-the-art results on benchmark datasets.

Bi-level Graph Learning Unveils Prognosis-Relevant Tumor Microenvironment Patterns in Breast Multiplexed Digital Pathology

This study introduces a bi-level graph model to identify prognostic tumor microenvironment patterns in breast cancer, validated across multiple cohorts.

Annotation-Free Deep Learning for Predicting Gene Mutations from Whole Slide Images of Acute Myeloid Leukemia

This paper introduces a deep learning model using multiple instance learning to predict gene mutations from whole slide images of AML without manual annotations, achieving high AUC scores for NPM1 and FLT3-ITD mutations.

PHARAOH: A collaborative crowdsourcing platform for PHenotyping And Regional Analysis Of Histology

PHARAOH enables collaborative histopathology analysis through a crowdsourcing platform that uses weakly supervised active learning to develop and share custom computer vision models for tissue image annotation.

Contrastive Learning for Omics-guided Whole-slide Visual Embedding Representation

CLOVER integrates multi-omics data with whole-slide images to generate molecular-aware slide representations, outperforming existing methods in cancer subtype classification and biomarker prediction.

A Deep Learning Framework for Predicting Prognostically Relevant Consensus Molecular Subtypes in HPV-Positive Cervical Squamous Cell Carcinoma from Routine Histology Images

This paper develops a deep learning framework to predict consensus molecular subtypes in HPV-positive cervical cancer from histology images, showing significant survival stratification and TME differences between subtypes.

Enhancing Breast Cancer Classification with Stacked Convolutional Neural Networks for Whole-Slide Image Analysis

This paper uses stacked convolutional neural networks to classify breast cancer from whole-slide images, achieving 90% accuracy in cancer detection but lower performance in distinguishing between benign, DCIS, and IDC lesions.

Deep Learning-Based Classification of Colorectal Cancer in Histopathology Images for Category Detection

This paper applies deep learning models to classify colorectal cancer categories in histopathology images, achieving high accuracy with ResNet and Swin Transformer models.

Histolytics: A Panoptic Spatial Analysis Framework for Interpretable Histopathology

Histolytics provides a framework for interpretable spatial analysis of histopathology images, enabling high-resolution characterization of tissue structures and features.

PathoEye: a deep learning framework for the whole-slide image analysis of skin tissue

PathoEye is a deep learning framework for analyzing whole-slide skin images, focusing on the dermis-epidermis junction to identify features of skin diseases and aging.

TLS_Finder: An algorithm for Identifying Tertiary Lymphoid Structures Using Immune Cell Spatial Coordinates

Histopathological Domain Adaptation with Generative Adversarial Networks - Bridging the Domain Gap Between Thyroid Cancer Histopathology Datasets

This paper uses GANs to generate synthetic thyroid histopathology images, improving model generalizability and performance on minority classes.

AxonFinder: Automated segmentation of tumor innervating neuronal fibers

AxonFinder segments axons in prostate tumor microenvironments using deep learning, showing correlation between axon density and cancer risk.

ecPath detects ecDNA in tumors from histopathology images

ecPath detects ecDNA in tumors using H&E images, achieving good performance in predicting ecDNA status and identifying poor prognoses in low grade glioma patients.

Deep Gaussian Process with Uncertainty Estimation for Microsatellite Instability and Immunotherapy Response Prediction Based on Histology

MSI-SEER uses deep Gaussian processes to predict microsatellite instability from histology images, achieving high accuracy in MSI prediction and ICI response prediction.

High-accuracy prediction of 5-year progression-free melanoma survival using spatial proteomics and few-shot learning on primary tissue

This paper predicts 5-year progression-free survival in melanoma using spatial proteomics and few-shot learning, achieving over 90% accuracy.