Latent Space Explorer: Visual Analytics for Multimodal Latent Space Exploration

TL;DR

Latent Space Explorer is a visual analytics tool designed to help medical experts explore and interpret multimodal latent representations from machine learning models, facilitating subgroup analysis and decoding of multimodal data.

Contribution

We developed Latent Space Explorer, a novel interactive visualization system that enables exploration and analysis of multimodal latent spaces in medical data.

Findings

Medical experts found the tool useful for subgroup analysis.

The system helped in understanding multimodal representations.

User feedback suggested directions for further development.

Abstract

Machine learning models built on training data with multiple modalities can reveal new insights that are not accessible through unimodal datasets. For example, cardiac magnetic resonance images (MRIs) and electrocardiograms (ECGs) are both known to capture useful information about subjects' cardiovascular health status. A multimodal machine learning model trained from large datasets can potentially predict the onset of heart-related diseases and provide novel medical insights about the cardiovascular system. Despite the potential benefits, it is difficult for medical experts to explore multimodal representation models without visual aids and to test the predictive performance of the models on various subpopulations. To address the challenges, we developed a visual analytics system called Latent Space Explorer. Latent Space Explorer provides interactive visualizations that enable users to explore the multimodal representation of subjects, define subgroups of interest, interactively decode data with different modalities with the selected subjects, and inspect the accuracy of the embedding in downstream prediction tasks. A user study was conducted with medical experts and their feedback provided useful insights into how Latent Space Explorer can help their analysis and possible new direction for further development in the medical domain.