Counterfactual Diffusion Models Provide Interpretable Explanations of Artificial Intelligence Models in Pathology
Cancer Research, vol. 86, pp. 4414–4433
Abstract
Deep learning can extract predictive and prognostic biomarkers from histopathology whole-slide images. However, explainable artificial intelligence approaches widely used in digital pathology, such as attention heatmaps and class activation mapping, provide limited insight into the image features associated with classifier outputs. In this study, we developed Morphing histoPathology Diffusion (MoPaDi), a framework for generating counterfactual explanations for histopathology images that help identify morphologic or stain-related features linked to model predictions. MoPaDi combined diffusion autoencoders with task-specific multiple instance learning classifiers to manipulate images and induce prediction shifts by modifying classifier-associated features. The framework was evaluated on multiple datasets spanning colorectal, breast, liver, and lung cancers, including tasks for tissue type, cancer subtype, and biomarker [microsatellite instability (MSI)] classification. MoPaDi generated perceptually realistic counterfactual histopathology images, enabling pathologists to identify morphologic features associated with changes in model predictions, complementing the conventional inspection of highly attended regions in digital pathology. In the MSI status prediction task, MoPaDi highlighted morphologic features linked to classifier predictions, including mucinous differentiation, altered glandular architecture, and lymphocytic infiltration, consistent with prior literature. Analyses separating stain-related from morphology-related components suggested that in this setting, prediction changes were predominantly associated with morphology-related rather than stain-related alterations. Overall, MoPaDi is a practical framework for counterfactual explanations in computational pathology that supports the evaluation of model-specific decision cues and hypothesis generation. SIGNIFICANCE: MoPaDi is a diffusion-based tool for counterfactual image generation in cancer histopathology that reveals features associated with deep learning classifier predictions and supports transparent auditing of computational models in biomedical research.
Authors 19
-
University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
-
University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
-
Universitätsklinikum Aachen · University of Pennsylvania · RWTH Aachen University
Affiliation as printed
2Department of Diagnostic and Interventional Radiology, University Hospital RWTH Aachen, Aachen, Germany
3Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
-
University of Augsburg · University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
5Bavarian Cancer Research Center (BZKF), Augsburg, Germany
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
University of Augsburg 4 Pathology, Faculty of Medicine, , Augsburg, Germany
-
University Medical Center of the Johannes Gutenberg University Mainz
Affiliation as printed
University Medical Center of the Johannes Gutenberg University Mainz 6 Institute of Pathology, , Mainz, Germany
-
University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
-
Philipps University of Marburg
Affiliation as printed
Philipps University of Marburg 7 Institute of Pathology, , Marburg, Germany
-
University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
-
University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
-
University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
-
Charité - Universitätsmedizin Berlin
Affiliation as printed
Charité – Universitätsmedizin Berlin 8 Department of Gastroenterology and Hepatology, Campus Charité Mitte, Campus Virchow Klinikum, , Berlin, Germany
-
Charité - Universitätsmedizin Berlin
Affiliation as printed
Charité – Universitätsmedizin Berlin 8 Department of Gastroenterology and Hepatology, Campus Charité Mitte, Campus Virchow Klinikum, , Berlin, Germany
-
Charité - Universitätsmedizin Berlin
Affiliation as printed
Charité – Universitätsmedizin Berlin 9 Department of Surgery, , Berlin, Germany
-
Charité - Universitätsmedizin Berlin
Affiliation as printed
Charité – Universitätsmedizin Berlin 9 Department of Surgery, , Berlin, Germany
-
Medical University of Vienna · Charité - Universitätsmedizin Berlin
Affiliation as printed
Charité – Universitätsmedizin Berlin 9 Department of Surgery, , Berlin, Germany
Medical University of Vienna 10 Division of Transplantation, Department of General Surgery, , Vienna, Austria
-
Universitätsklinikum Aachen · RWTH Aachen University
Affiliation as printed
11Department of Internal Medicine III, Gastroenterology, Metabolic Diseases and Intensive Care, University Hospital RWTH Aachen, Aachen, Germany
-
University of Chicago · Chan Zuckerberg Biohub Chicago
Affiliation as printed
13Chan Zuckerberg Biohub Chicago LLC, Chicago, Illinois
University of Chicago 12 Section of Hematology/Oncology, Department of Medicine, , Chicago, Illinois
-
Universitätsklinikum Aachen · RWTH Aachen University
Affiliation as printed
2Department of Diagnostic and Interventional Radiology, University Hospital RWTH Aachen, Aachen, Germany
-
University of Leeds · University Hospital Heidelberg · National Center for Tumor Diseases · University Hospital Carl Gustav Carus · Else Kröner Fresenius Center for Digital Health · Technische Universität Dresden
Affiliation as printed
15Medical Oncology, National Center for Tumor Diseases (NCT), University Hospital Heidelberg, Heidelberg, Germany
16Pathology and Data Analytics, Leeds Institute of Medical Research at St James’s, University of Leeds, Leeds, United Kingdom
TUD Dresden University of Technology 1 Else Kröner Fresenius Center for Digital Health (EKFZ), Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
TUD Dresden University of Technology 14 Department of Medicine I, Faculty of Medicine and University Hospital Carl Gustav Carus, , Dresden, Germany
Cited by 0 stored of 0
No patents citing this paper on Lens.org (checked 2026-10-06).