Two-Stage Representation Learning for Analyzing Movement Behavior Dynamics in People Living with Dementia

TL;DR

This work tackles the challenge of extracting meaningful patterns from high-frequency home activity data of people living with dementia. It introduces a two-stage, self-supervised pipeline that converts temporal activity into text, encodes it with a pretrained language model, reduces to a 2D space, and then applies PageRank on a latent-state transition graph to produce compact state vectors. The resulting low-rank representations enable clustering, transition analysis, and improved prediction of cognitive status metrics such as MMSE and ADAS-Cog, with state-based features outperforming baselines in multiple predictive models. The approach demonstrates potential for interpretable, scalable remote monitoring and personalized care planning, while acknowledging the need for larger-scale validation and integration into clinical workflows.

Abstract

In remote healthcare monitoring, time series representation learning reveals critical patient behavior patterns from high-frequency data. This study analyzes home activity data from individuals living with dementia by proposing a two-stage, self-supervised learning approach tailored to uncover low-rank structures. The first stage converts time-series activities into text sequences encoded by a pre-trained language model, providing a rich, high-dimensional latent state space using a PageRank-based method. This PageRank vector captures latent state transitions, effectively compressing complex behaviour data into a succinct form that enhances interpretability. This low-rank representation not only enhances model interpretability but also facilitates clustering and transition analysis, revealing key behavioral patterns correlated with clinicalmetrics such as MMSE and ADAS-COG scores. Our findings demonstrate the framework's potential in supporting cognitive status prediction, personalized care interventions, and large-scale health monitoring.

Figures (30)

• Figure 1: Example Home layout with IOT sensors for monitoring behavioral patterns of People Living with Dementia
• Figure 2: Flowchart of data preprocessing. The figure illustrates the monitoring data for a single participant over the course of one day. The left graph displays the raw, unprocessed measurements. In the middle graph, the data is rectified into 20-minute intervals, where periods of inactivity are labeled as "nowhere." Within each window, the most frequent location, excluding "nowhere," is identified and recorded. The right graph presents the corresponding text strings, which are formatted for interpretation by the language model.
• Figure 3: Flowchart of the representation algorithm.
• Figure 4: T-SNE for embedded daily movement strings in the test set
• Figure 5: Multi-period participant deep state vector and similarity