Track genomics, neuroscience and molecular biology.
Search papers, labs, and topics across Lattice.
CPathOGen is a conditional latent diffusion model designed to generate spatially and morphologically controlled H&E counterfactuals for computational pathology, addressing the challenge of creating plausible tissue perturbations for model interpretability. By conditioning on cellular maps and using FiLM layers for morphology/appearance modulation, it produces visually realistic counterfactuals with controlled biological and stain variations, validated by improved FID/KID scores and monotonic feature tracking. This framework enables robust probing of pathology models' sensitivity to specific biological features versus nuisance variations, quantified by prediction-flip rates and a novel Biology-Nuisance Sensitivity Ratio.
Generating H&E counterfactuals with explicit control over cellular spatial organization and nuclear morphology reveals pathology models' unexpected sensitivity to subtle stain variations.
Computational pathology models infer biologically and clinically meaningful outcomes from histology, but their predictions are shaped by complex, intertwined tissue signals whose roles are important to understand. Common pixel- and feature-space perturbations can produce implausible tissue, making model responses difficult to interpret. We introduce CPathOGen, a conditional latent-diffusion framework for generating paired H\&E counterfactuals with explicit controls over cellular spatial organization, nuclear morphology, and stain appearance. Cellular maps condition spatial structure through a spatial encoder, while a morphology/appearance vector modulates denoising through blockwise feature-wise linear modulation (FiLM). On held-out H\&E tiles, CPathOGen generates visually plausible tissue with improved distributional agreement after spatially guided selection, as reflected by lower Fr\'echet Inception Distance (FID) and Kernel Inception Distance (KID); generated cells track requested abundance and position, and measured morphology and color vary monotonically with their controls. We use these verified interventions to probe pathology encoders with endpoint heads, task-specific classifiers, and survival models. Responses are quantified using total variation distance and prediction-flip rate. We further introduce the Biology-Nuisance Sensitivity Ratio, a metric that contrasts model sensitivity to biologically motivated morphology and spatial factors with sensitivity to non-biological, stain-related nuisance variation. CPathOGen provides a practical, fidelity-audited framework for evaluating robustness and controlled feature sensitivity in computational pathology model. \href{https://github.com/a12dongithub/PathOGen}{GitHub} and \href{https://huggingface.co/a12donhf/CPathOGen}{Hugging~Face}.