A public–private research project
Using machine learning and mobile health data to predict and prevent relapse and treatment dropout in alcohol use disorder and hazardous alcohol consumption.
Affiliated with
The work
Use of existing real-world breathalyser data from adults with Alcohol Use Disorder who used a breathalyser as part of their one-year treatment. We assess the quality and completeness of the available data and explore how adherence impacts what we can learn from remote breath sampling, then develop and compare initial machine-learning models for drinking events, treatment dropout and relapse.
Status: completedInterviews with former AUD patients, clinicians, academic experts and experts by experience identify facilitators and barriers: usability, burden, acceptability, trust and implementation needs.
Status: in progressA prospective study in adults with ADHD and hazardous alcohol use, testing whether drinking events can be predicted from mobile health data. We combine ecological momentary assessment, passively collected smartphone signals and validated questionnaires, with PEth blood samples to objectively validate self-reported alcohol use.
Status: plannedAn observational study testing whether the digital biomarker with the final set of features generalises to a different patient population, and whether its predictions are clinically meaningful to patients and clinicians.
Status: plannedOur publications
An overview of how mHealth technologies are used to monitor and support people with Alcohol Use Disorder, synthesising evidence from smartphones, breathalyzers and wearable sensors tracking alcohol use in near-real time.
Read hereNearly 80,000 daily breathalyzer observations from 246 patients. A multilayer LSTM network performed well at predicting drinking events, while ensemble models were more effective for dropout prediction.
Read hereMini-symposium
Translating research into responsible practice.
Organised by BRISA and Amsterdam UMC, followed by a networking borrel.
Scan to register
Get in contact
For questions about the research, collaboration, or the September symposium.
