← Toutes les offres
TU

Research Associate in Machine Learning for Digital Phenotyping

The University of Manchester

Manchester
CDD
Temps plein
Health & Medical
Publié le 20 novembre 2026

Description du poste

Job reference: BMH-032692 Salary: £38,448 - £46,970 per annum depending on experience Faculty/Organisational Unit: Biology, Medicine Health Location: Oxford Road Employment type: Fixed Term Division/Team: Division of Informatics, Imaging & Data Science Hours Per Week: 1FTE Closing date (DD/MM/YYYY): 30/11/2026 Contract Duration: 8 months School/Directorate: School of Health Sciences We are seeking a Research Associate in Machine Learning and Digital Phenotyping to work across multiple projects. The first project is “CONNECT: Digital markers to predict psychosis relapse”. This project will recruit individuals with psychosis and use smart phone apps to collect passive and active data using a prospective observational cohort study design. We will use this data to develop and validate a personalised risk prediction algorithm for relapse. The second project is the Mental Health Mission, a £42m UK Government investment into new infrastructure for mental health. The data and digital theme is focussed on developing new digital phenotypes from multiple forms of data including electronic health records, smartphones and wearables. The aim is to develop transdiagnostic digital biomarkers from patient generated health data. The postholder’s main duty will be to provide machine learning in the study, with responsibility for: Development, and validation, of a risk prediction model for relapse. Using data from the prospective study, you will develop/train a risk prediction model that can be used to identify patients at risk of relapse. This will use data from a variety of sources including symptoms recorded in real-time through a mobile phone app, and passively collected data such as geolocation. The model will be dynamic (updating predictions as patients record new data). An experimental approach will be taken to explore different predictors and machine learning methods. Development of digital phenotypes. You will use emerging data from CONNECT to identify digital phenotypes and different clusters of phenotypes and how they predict relapse. By linking the CONNECT data to electronic health records you will develop extended digital phenotypes. Develop novel approaches to standardising digital biomarkers from patient generated data. An objective of the group is to develop repeatable approaches to extracting digital phenotypes form patient generated data where there is heterogeneity across devices used to generate and collect data. You must have a PhD (or equivalent) in artificial intelligence, and be developing your publication record. You must have specific skills and expertise in applied machine learning to healthcare problems Enquiries about the vacancy, shortlisting and interviews: Name: Professor John Ainsworth Email: john.ainsworth@macnhester.ac.uk General enquiries and administrative support recruitmentservices.people@manchester.ac.uk Technical and job portal support https://knowledge.jobtrain.co.uk/jobseeker [knowledge.jobtrain.co.uk] Applications close at midnight on the closing date.