Visual-Conversational Interface for Evidence-Based Explanation of Diabetes Risk Prediction

TL;DR

This paper introduces a visual-conversational system for explaining diabetes risk predictions, combining interactive visualizations, scientific evidence grounding, and hybrid language models to improve healthcare professionals' understanding and trust.

Contribution

It presents an integrated decision support system that enhances AI explanations with scientific grounding and hybrid prompt handling, addressing limitations of existing systems.

Findings

Healthcare professionals gained clearer understanding of model assessments.

Scientific evidence grounding calibrated trust in AI decisions.

Participants found the system supported risk evaluation and recommendations.

Abstract

Healthcare professionals need effective ways to use, understand, and validate AI-driven clinical decision support systems. Existing systems face two key limitations: complex visualizations and a lack of grounding in scientific evidence. We present an integrated decision support system that combines interactive visualizations with a conversational agent to explain diabetes risk assessments. We propose a hybrid prompt handling approach combining fine-tuned language models for analytical queries with general Large Language Models (LLMs) for broader medical questions, a methodology for grounding AI explanations in scientific evidence, and a feature range analysis technique to support deeper understanding of feature contributions. We conducted a mixed-methods study with 30 healthcare professionals and found that the conversational interactions helped healthcare professionals build a clear understanding of model assessments, while the integration of scientific evidence calibrated trust in the system's decisions. Most participants reported that the system supported both patient risk evaluation and recommendation.