Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PhysioLLM: Supporting Personalized Health Insights with Wearables and Large Language Models

Cathy Mengying Fang, Valdemar Danry, Nathan Whitmore, Andria Bao, Andrew Hutchison, Cayden Pierce, Pattie Maes

TL;DR

PhysioLLM is presented, an interactive system that leverages large language models (LLMs) to provide personalized health understanding and exploration by integrating physiological data from wearables with contextual information and offers a comprehensive statistical analysis component.

Abstract

We present PhysioLLM, an interactive system that leverages large language models (LLMs) to provide personalized health understanding and exploration by integrating physiological data from wearables with contextual information. Unlike commercial health apps for wearables, our system offers a comprehensive statistical analysis component that discovers correlations and trends in user data, allowing users to ask questions in natural language and receive generated personalized insights, and guides them to develop actionable goals. As a case study, we focus on improving sleep quality, given its measurability through physiological data and its importance to general well-being. Through a user study with 24 Fitbit watch users, we demonstrate that PhysioLLM outperforms both the Fitbit App alone and a generic LLM chatbot in facilitating a deeper, personalized understanding of health data and supporting actionable steps toward personal health goals.

PhysioLLM: Supporting Personalized Health Insights with Wearables and Large Language Models

TL;DR

PhysioLLM is presented, an interactive system that leverages large language models (LLMs) to provide personalized health understanding and exploration by integrating physiological data from wearables with contextual information and offers a comprehensive statistical analysis component.

Abstract

We present PhysioLLM, an interactive system that leverages large language models (LLMs) to provide personalized health understanding and exploration by integrating physiological data from wearables with contextual information. Unlike commercial health apps for wearables, our system offers a comprehensive statistical analysis component that discovers correlations and trends in user data, allowing users to ask questions in natural language and receive generated personalized insights, and guides them to develop actionable goals. As a case study, we focus on improving sleep quality, given its measurability through physiological data and its importance to general well-being. Through a user study with 24 Fitbit watch users, we demonstrate that PhysioLLM outperforms both the Fitbit App alone and a generic LLM chatbot in facilitating a deeper, personalized understanding of health data and supporting actionable steps toward personal health goals.
Paper Structure (23 sections, 4 figures, 1 table)

This paper contains 23 sections, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. LLMs For Health Prediction
  4. LLM-based Data Analysis
  5. LLM for Personal Health Insight Generation
  6. Motivation and Design Goals
  7. PhysioLLM Architecture and Implementation
  8. Data Preparation
  9. Data Filtering and Alignment
  10. Generation of Summary, Trends and Correlations
  11. User Modeler and Insight Generation
  12. Conversational Interface Design
  13. User Study
  14. Procedure
  15. Conditions
  16. ...and 8 more sections

Figures (4)

  • Figure 1: Overview of the PhysioLLM system with an example conversation.
  • Figure 2: The steps taken to summarize the data and generate insights from the data before the information is passed to the conversation LLM.
  • Figure 3: The study protocol.
  • Figure 4: Barplots of Likert-scale ratings. Higher ratings are better. Error bar: SE. *:p$<$.05,**:p$<$.01 $\Delta$: difference between pre- and post-survey.