Generating Counterfactual Trajectories with Latent Diffusion Models for Concept Discovery

2024-04-16

International Conference on Pattern Recognition

AI Summary

To address the limited interpretability of black-box medical AI models and their inability to support medical knowledge discovery, this paper proposes the CDCT framework—a three-stage unsupervised disentangled concept discovery method. It introduces latent diffusion models (LDMs) for the first time into unsupervised concept discovery, generating high-fidelity counterfactual trajectories to expose implicit decision logic. The approach integrates VAE-based latent-space modeling, latent-space concept search, and counterfactual trajectory generation, balancing computational efficiency and semantic fidelity. Evaluated on the largest publicly available skin lesion dataset, CDCT successfully identifies clinically relevant biomarkers and uncovers hidden model biases. Generated counterfactual images achieve superior Fréchet Inception Distance (FID) compared to state-of-the-art methods, while reducing computational overhead by 12×. This work establishes a scalable, concept-level explanation paradigm for trustworthy medical AI.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Calibration & Uncertainty Quantification
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Trustworthiness is a major prerequisite for the safe application of opaque deep learning models in high-stakes domains like medicine. Understanding the decision-making process not only contributes to fostering trust but might also reveal previously unknown decision criteria of complex models that could advance the state of medical research. The discovery of decision-relevant concepts from black box models is a particularly challenging task. This study proposes Concept Discovery through Latent Diffusion-based Counterfactual Trajectories (CDCT), a novel three-step framework for concept discovery leveraging the superior image synthesis capabilities of diffusion models. In the first step, CDCT uses a Latent Diffusion Model (LDM) to generate a counterfactual trajectory dataset. This dataset is used to derive a disentangled representation of classification-relevant concepts using a Variational Autoencoder (VAE). Finally, a search algorithm is applied to identify relevant concepts in the disentangled latent space. The application of CDCT to a classifier trained on the largest public skin lesion dataset revealed not only the presence of several biases but also meaningful biomarkers. Moreover, the counterfactuals generated within CDCT show better FID scores than those produced by a previously established state-of-the-art method, while being 12 times more resource-efficient. Unsupervised concept discovery holds great potential for the application of trustworthy AI and the further development of human knowledge in various domains. CDCT represents a further step in this direction.

Problem

• Interpretable Machine Learning
• Medical Decision-making
• Black-box Models

Innovation

• CDCT
• LDM-VAE Integration
• Efficient Resource Consumption