Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Care for the Mind Amid Chronic Diseases: An Interpretable AI Approach Using IoT

Jiaheng Xie, Xiaohang Zhao, Xiang Liu, Xiao Fang

TL;DR

This work targets depression detection in chronic disease patients using motion sensor data from wearables. It introduces TempPNet, an interpretable model that learns two types of prototypes: symptom prototypes capturing depression-related gait patterns and trend prototypes modeling temporal progression of symptoms, enabling the classifier to both predict depression and visualize its rationale. Empirical results on NHANES and mPower show TempPNet outperforms strong baselines and existing MTSC/prototype methods, with ablations confirming the importance of temporal prototypes and time alignment. Human evaluations with a large user study and a medical expert panel demonstrate improved interpretability and clinical usefulness, highlighting TempPNet's potential for real-time, collaborative care via mobile health IoT systems.

Abstract

Health sensing for chronic disease management creates immense benefits for social welfare. Existing health sensing studies primarily focus on the prediction of physical chronic diseases. Depression, a widespread complication of chronic diseases, is however understudied. We draw on the medical literature to support depression detection using motion sensor data. To connect humans in this decision-making, safeguard trust, and ensure algorithm transparency, we develop an interpretable deep learning model: Temporal Prototype Network (TempPNet). TempPNet is built upon the emergent prototype learning models. To accommodate the temporal characteristic of sensor data and the progressive property of depression, TempPNet differs from existing prototype learning models in its capability of capturing temporal progressions of prototypes. Extensive empirical analyses using real-world motion sensor data show that TempPNet outperforms state-of-the-art benchmarks in depression detection. Moreover, TempPNet interprets its decision by visualizing the temporal progression of depression and its corresponding symptoms detected from sensor data. We further employ a user study and a medical expert panel to demonstrate its superiority over the benchmarks in interpretability. This study offers an algorithmic solution for impactful social good -- collaborative care of chronic diseases and depression in health sensing. Methodologically, it contributes to extant literature with a novel interpretable deep learning model for depression detection from sensor data. Patients, doctors, and caregivers can deploy our model on mobile devices to monitor patients' depression risks in real-time. Our model's interpretability also allows human experts to participate in the decision-making by reviewing the interpretation and making informed interventions.

Care for the Mind Amid Chronic Diseases: An Interpretable AI Approach Using IoT

TL;DR

This work targets depression detection in chronic disease patients using motion sensor data from wearables. It introduces TempPNet, an interpretable model that learns two types of prototypes: symptom prototypes capturing depression-related gait patterns and trend prototypes modeling temporal progression of symptoms, enabling the classifier to both predict depression and visualize its rationale. Empirical results on NHANES and mPower show TempPNet outperforms strong baselines and existing MTSC/prototype methods, with ablations confirming the importance of temporal prototypes and time alignment. Human evaluations with a large user study and a medical expert panel demonstrate improved interpretability and clinical usefulness, highlighting TempPNet's potential for real-time, collaborative care via mobile health IoT systems.

Abstract

Health sensing for chronic disease management creates immense benefits for social welfare. Existing health sensing studies primarily focus on the prediction of physical chronic diseases. Depression, a widespread complication of chronic diseases, is however understudied. We draw on the medical literature to support depression detection using motion sensor data. To connect humans in this decision-making, safeguard trust, and ensure algorithm transparency, we develop an interpretable deep learning model: Temporal Prototype Network (TempPNet). TempPNet is built upon the emergent prototype learning models. To accommodate the temporal characteristic of sensor data and the progressive property of depression, TempPNet differs from existing prototype learning models in its capability of capturing temporal progressions of prototypes. Extensive empirical analyses using real-world motion sensor data show that TempPNet outperforms state-of-the-art benchmarks in depression detection. Moreover, TempPNet interprets its decision by visualizing the temporal progression of depression and its corresponding symptoms detected from sensor data. We further employ a user study and a medical expert panel to demonstrate its superiority over the benchmarks in interpretability. This study offers an algorithmic solution for impactful social good -- collaborative care of chronic diseases and depression in health sensing. Methodologically, it contributes to extant literature with a novel interpretable deep learning model for depression detection from sensor data. Patients, doctors, and caregivers can deploy our model on mobile devices to monitor patients' depression risks in real-time. Our model's interpretability also allows human experts to participate in the decision-making by reviewing the interpretation and making informed interventions.
Paper Structure (43 sections, 21 equations, 15 figures, 25 tables, 2 algorithms)

This paper contains 43 sections, 21 equations, 15 figures, 25 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Health IS and Chronic Disease Management
  4. Wearable Sensor Technology and Depression
  5. Interpretable Machine Learning and Prototype Learning
  6. MTSC, Prototype Learning for MTSC, and Time Series Deep Learning
  7. The TempPNet Approach
  8. Problem Setup
  9. The Symptom Prototype and Symptom Progression Matrix
  10. Detecting Prototypical Depression Symptom Progression Trends
  11. The Definition of Trend Prototypes:
  12. Detecting the Existing Strength of a Trend Prototype:
  13. Superior Properties of the Trend Prototype:
  14. Learning Objective
  15. Prototype Visualization
  16. ...and 28 more sections

Figures (15)

  • Figure 1: Temporal Progression of Depression bockting_lifetime_2015
  • Figure 2: Our Method v.s. Prototype Learning for MTSC
  • Figure 3: TempPNet Architecture
  • Figure 4: Latent Trend Starting Time
  • Figure 5: Trend Prototypes
  • ...and 10 more figures