SHADE-AD: An LLM-Based Framework for Synthesizing Activity Data of Alzheimer's Patients

2025-03-03

AI Summary

A critical bottleneck in Alzheimer’s disease (AD) intelligent monitoring is the severe scarcity of real-world activity videos capturing AD-specific behaviors. Method: We propose the first large language model (LLM)-driven, three-stage synthetic framework for AD-specific behavioral modeling and cross-scenario activity generalization. Integrating multi-source data, the framework employs progressive training, action-semantic alignment, and kinematic constraints to generate high-fidelity, privacy-preserving AD activity video datasets. Contribution/Results: Our approach preserves data authenticity while drastically reducing acquisition costs. Experiments show that human activity recognition (HAR) models trained on synthetic data achieve 79.69% accuracy—surpassing prior baselines—and kinematic validation confirms strong fidelity to real-world motion patterns. This work establishes a scalable, reproducible data infrastructure for AD digital phenotyping and intelligent monitoring system development.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Computer Vision: Video Understanding & Activity Analysis

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Alzheimer's Disease (AD) has become an increasingly critical global health concern, which necessitates effective monitoring solutions in smart health applications. However, the development of such solutions is significantly hindered by the scarcity of AD-specific activity datasets. To address this challenge, we propose SHADE-AD, a Large Language Model (LLM) framework for Synthesizing Human Activity Datasets Embedded with AD features. Leveraging both public datasets and our own collected data from 99 AD patients, SHADE-AD synthesizes human activity videos that specifically represent AD-related behaviors. By employing a three-stage training mechanism, it broadens the range of activities beyond those collected from limited deployment settings. We conducted comprehensive evaluations of the generated dataset, demonstrating significant improvements in downstream tasks such as Human Activity Recognition (HAR) detection, with enhancements of up to 79.69%. Detailed motion metrics between real and synthetic data show strong alignment, validating the realism and utility of the synthesized dataset. These results underscore SHADE-AD's potential to advance smart health applications by providing a cost-effective, privacy-preserving solution for AD monitoring.

Problem

• Addresses scarcity of Alzheimer's-specific activity datasets.
• Synthesizes AD-related human activity videos using LLM.
• Improves Human Activity Recognition detection by 79.69%.

Innovation

• LLM-based framework synthesizes AD-specific activity data
• Three-stage training mechanism enhances activity range
• Synthetic data improves Human Activity Recognition accuracy