Perfect Fit

Targeting key risk factors for cardiovascular disease in at-risk individuals using a personalized and adaptive approach

Smoking tobacco and physical inactivity are key preventable risk factors of cardiovascular disease (CVD). Perfect Fit combines big-data science, sensor technology, and personalized real-time feedback for individuals to achieve and maintain abstinence from smoking and sufficient physical activity (PA; in gyms/daily life), prevent CVD, facilitate wellbeing and reduce healthcare costs. The approach is designed to serve disadvantaged groups, where smoking and physical inactivity are most prevalent. Once developed, the approach can be extended to include other risk factors and systems.

The project develops tailored, evidence-based, near real-time computer coaching for quitting smoking and enhancing PA. For every individual, a personal model is designed which generates personalized recommendations based on high-quality existing and newly collected data, and adapts to changing circumstances/progress (similar to a TomTom navigation system), using machine learning techniques and incorporating domain-specific expert knowledge (e.g. health behaviour change strategies). Virtual coaches (VCs) communicate advice in a motivating way that fits individuals’ persuasive communicationstyles.

Key objectives are: To examine how different types of (senso-)data can be used to provide personalized advice; which adaptivity is needed to create a robust, safe, and effective interaction between individuals and machine (VC); how advanced data science methods can be developed and embedded in current smoking cessation and PA coaching practice; and how measurement modalities, feedback and communication affect individuals’ smoking status and PA.

The project addresses challenges regarding small and big-data quality and usability; linkage of data sources; making data meaningful for individuals; and responsible data science.

So far the Netherlands eScience Center has been involved in: The creation of a patient journey that describes the interaction of the patient with the virtual coach throughout the different stages of the intervention, and the design of the system architecture that will fuel the virtual coach.

Currently we are working on building the backend system for the virtual coach, see our Github organisation for more details on the software that we are building: https://github.com/PerfectFit-project .

Participating organisations

Delft University of Technology
Leiden University Medical Center
Life Sciences
Life Sciences
Netherlands eScience Center
University of Twente

Output

Team

NC
N.H. Chavannes
Principal investigator
Leiden University Medical Center
Cunliang Geng
eScience Research Engineer
Netherlands eScience Center
Sven van der Burg
Sven van der Burg
eScience Research Engineer
Netherlands eScience Center
Djura Smits
Djura Smits
Robin Richardson
Robin Richardson
Lars Ridder
Lars Ridder
eScience Coordinator
Netherlands eScience Center

Related projects

AdVantage

Improving the vantage6 federated learning ecosystem

Updated 6 months ago
In progress

CARRIER

Coronary artery disease: risk estimations and interventions for prevention and Early detection –...

Updated 12 months ago
In progress

STRAP

Self tracking for prevention and diagnosis of heart disease

Updated 12 months ago
In progress

MyDigiTwin

Using big data to put a cardiovascular digital twin into the hands of people

Updated 12 months ago
In progress

FEDMix

Fusible evolutionary deep neural network mixture learning from distributed data for robust medical...

Updated 11 months ago
Finished

Genetics of sleep patterns

Detecting human sleep from wearable accelerometer data without the aid of sleep diaries

Updated 1 month ago
Finished

Diagnosis of active epilepsy in resource-poor setting

Prediction models based on EEG characteristics

Updated 11 months ago
Finished