Remote Monitoring of Food-Intake of Older Cardiometabolic Patients in Home Environment

Research Line: Health @ Home / Seed Call: i4PH April 2024

The world is currently experiencing unprecedented population aging, with the proportion of people aged 65 and older projected to reach 22% of the total population over the next four decades. This demographic shift is expected to place a significant burden on healthcare systems, which is being addressed through increased self-management and home-based care. Dietary intake plays a crucial role in maintaining health, especially among the elderly, whose nutrient intake often falls short of recommended levels, which may affect their cognitive and functional states.

Recent research has focused on leveraging advanced technologies to monitor food intake. These techniques, ranging from traditional methods like dietary diaries to modern smartphone applications backed by recommendation systems, aim to provide the most accurate monitoring results. Implementing these technologies at home is vital for improving the quality of life for the elderly and reducing healthcare costs.

However, this demographic has specific requirements for technology use, leading to low adherence rates and reducing the long-term efficacy of such interventions. Also, the feasibility of such technology remains to be fully assessed specifically for older cardiometabolic patients as well as for food intake monitoring.

Objectives and Route to Impact

Addressing the challenges of dietary monitoring in the elderly could be achieved by leveraging affordable, commodity-level wearables such as smartwatches, which come equipped with various sensors and offer full software customisation. These devices enable both in-situ sensor data collection (objective) and self-reporting (subjective).

However, wrist-worn smartwatches alone are insufficient for analysing food composition accurately. Therefore, CardioWatch proposes investigating and assessing the feasibility of complementary, commercially available sensors such as vitamin trackers, smart toilet seats, smart bands, and continuous glucose monitors. These will be used in conjunction with equipment from previously approved projects within our team, including smart scales, Experiencer software, a customisable experience sampling method (ESM) tool (TU/e), and chew detection technology (WUR).

The objectives are to evaluate the unobtrusiveness, accuracy, reliability, and acceptance of these commercially available and pre-existing sensors and monitoring devices among older patients.

To achieve this, three key questions have been formulated:

• What are the feasibility criteria and risk factors associated with our technology combined with AI?
• What is the acceptance level and adherence of older cardiometabolic patients to our technological solution at home?
• What outlines can aid in constructing the specifications for developing a complete preventive remote monitoring and intervention platform?

Methods

The research method involves conducting experiments in a home environment with a cohort of older cardiometabolic patients recruited from WUR, UU, and UMCU. The primary tools for data collection and monitoring include smartwatches and the customisable Experiencer software, which facilitate both sensory data collection and self-reporting.

In addition to the smartwatches, complementary sensors are employed to provide ‘food consumption’-related sensory data. This combination of tools aims to provide a comprehensive and unobtrusive monitoring system for assessing body composition and dietary intake.

The methodology includes the following steps:

• Recruitment and Setup: Recruit older cardiometabolic patients and set up the monitoring devices in their homes.
• Data Collection: collect sensory data and self-reported information.
• Surveys and Interviews: gather qualitative data on participants’ experiences and acceptance of the technology.
• Data Analysis: Analyse the collected data to identify patterns and build machine-learning models.

By integrating these methods, CardioWatch aims to evaluate the feasibility, accuracy, reliability, and acceptance of our technological solution among older cardiometabolic patients, ultimately paving the way for future large-scale studies.

Deliverables

The research aims to develop an open-source, GDPR-compliant self-monitoring system using smartwatches, sensors, and customizable software for precise dietary and body composition data. Participants will have full data control. The findings will support grant applications for larger studies with low-risk, non-intrusive sensors. As data accumulates and AI models improve, user interaction will become less intrusive, boosting adherence to digital health interventions. The project will also explore federated and meta-learning techniques with wearable technologies for older cardiometabolic patients, enhancing their quality of life and preventive healthcare outcomes.

Contribution to cross-EWUU Collaboration

Fundamental knowledge from the nutritional and biological sciences is important, but it is also crucial to have clinical and psychological viewpoints, as well as practical applications that can be put into use in day-to-day practice. Experts in the fields of clinical practice, computer science, psychology, digital intervention, and cardio-vascular and nutrition science make up our proposed team.

Our areas of expertise complement each other and can be used in conjunction with AI for preventive health. TU/e (Dr. Van Gorp, Dr. Liang, and MSc. Khanshan) is key to the technology aspect with relevant expertise in the field of software engineering, human-computer interaction, wearable sensors, machine learning, data mining, and digital interventions. WUR (Prof. Dr. Geleijnse), and UMCU (Prof. Dr. ir. van der Schouw) ensure knowledge of lifestyle and specifically nutrition and health within the context of the older cardiometabolic patients at home. UU (Dr. Keizer) provides insights and expertise as to how the home monitoring environment is perceived and experienced from a psychological perspective.

Team

• Dr. Pieter Van Gorp – TU/e Industrial Engineering
• Dr. Rong-Hao Liang – TU/e Industrial Design
• Prof. Dr. J. Marianne Geleijnse – WUR Human Nutrition and Health
• Prof. Dr. ir. Yvonne T. van der Schouw – UMCU Department of Epidemiology
• Dr. Anouk Keizer – UU Experimental Psychology
• MSc. Alireza Khanshan – TU/e Industrial Design