# Digital biomarkers for early agitation detection in dementia: a scoping review of emerging wearable and smart technologies for personalized care

## TL;DR

This review explores how wearable technologies can detect agitation in dementia patients early, supporting personalized care through real-time monitoring.

## Contribution

The study provides a scoping review of emerging wearable and smart technologies for detecting agitation in dementia, emphasizing personalized care and human-centered design.

## Key findings

- Multimodal wearable sensors and machine learning models can reliably detect agitation symptoms in dementia patients.
- Digital phenotyping is a promising method for capturing complex behavioral patterns associated with agitation.
- User-centered design is crucial for the adoption and long-term use of these technologies.

## Abstract

Agitation is a common and burdensome symptom in people with dementia, particularly when compounded by impaired communication, making early detection and effective management difficult. Wearable sensor technologies may offer a promising avenue for supporting real-time behavioral monitoring and personalized care in this context.

This study aims to examine the current clinical and technological capabilities of wearable sensor systems for detecting and managing agitation in persons with dementia, to assess whether these technologies can effectively support personalized care. Additionally, it seeks to identify key challenges and opportunities in applying human-centered design principles and tailored interventions to improve outcomes for both patients and caregivers.

We conducted a scoping literature review, registered on OSF1 and guided by PRISMA-ScR guidelines. Five databases—Google Scholar, Scopus, PubMed, PsycINFO, and ACM Digital Library—were searched for English-language peer-reviewed studies published between 2016 and early 2025. From an initial pool of 798 articles, a multi-phase screening process led to a final inclusion of 13 studies that met predefined criteria.

The reviewed studies demonstrated that wearable sensors, particularly those employing multimodal data and personalized machine learning models, enable reliable detection of agitation symptoms and support timely, tailored interventions. The concept of digital phenotyping emerged as a promising approach for capturing complex behavioral signatures, while user-centered design was identified as essential for adoption and long-term compliance.

The evidence identified in this scoping review indicates that wearable and multimodal sensor technologies may offer promising approaches for monitoring agitation in dementia, while acknowledging that the research remains in early stages. We recommend future research focus on large-scale, longitudinal validation and the expansion of these tools to other populations with communication challenges, such as individuals with autism spectrum disorder or traumatic brain injury.

https://doi.org/10.17605/OSF.IO/DNHYM.