Our group member Xianjing Liu and colleagues have published a new article in Medical Image Analysis: AI-based association analysis for medial imaging using latent-space geometric confounder correction.
In AI-based medical image analysis, it is known to be difficult to derive clinically or epidemiologically significant insights from AI-generated results, due to the complexity of visualizing non-linear modeling in AI models (known as the black box issue), and the lack of control over confounding variables (which can lead to misleading findings).
In this study, a novel methodological pipeline is developed to address these challenges. We introduce geometric insights of confounders into the latent space, and propose a correlation-based loss function that performs confounder correction via vector orthogonalization. These simple yet effective designs enable semantic feature visualization in confounder-free AI prediction models, and easily handle multiple confounders.
The proposed method is demonstrated using a synthetic dataset, and further applied to two real medical imaging applications of brain imaging and 3D facial shape.
Available online: