PhD Project Proposal: Intelligent AI System for Predicting Depression Relapse in Young People using Digital Phenotypes

Key information

Start date: 2 March 2026

Duration: 4 years of registration, 3 years of stipend
Academic entry requirements: at least a high 2.1 bachelor’s degree in Neuroscience, Psychology or a related field, or on track to be awarded such a degree before 1 March 2026
Eligibility: the studentship is open to home/UK candidates

Deadline for applications: 30 November 2025

About the project

Research Question

How can smartphone and social media use, music listening and health data (digital phenotypes – DPs) be used for identification and early intervention of emerging mood problems in young people in higher education (YPHE)?

Background

DPs can improve clinical assessment and predict symptom exacerbation in patients with depression and anxiety. YPHE have higher risk of anxiety and depressive symptoms than their peers. Despite promising results for DPs, there is resistance due to data privacy issues. Previous work has (a) focused on older adults despite young people at greatest risk of depression, (b) been unable to solve data privacy issues, and (c) used invasive and expensive methods. Our DPs are universally available, free, non-invasive and have no burden on participants.

Aims & Objectives

To work closely with a student advisory group to co-develop and validate a privacy preserving AI tool that uses DP to predict anxiety and depressive symptom exacerbation in YPHE and test its utility for early identification.

Methods

Using a mixed-methods approach, we will use qualitatively explore students’ understanding of mental health, AI and DPs (work package 1), work with a PPI group address these issues (Work Package 2), develop AI pipelines based on feedback and carry out a longitudinal study with students to test the validity of our AI (WP3).

Analysis

Alongside our qualitative work, we will co-develop AI models to predict worsening of mood using digital phenotypes. We will Deep Learning models (e.g. Long Short-Term Memory (LSTM)) to model DPs over time, incorporating patient views as part of our AI, and deploy using a Federated Learning approach to ensure privacy.

Impact

This project will result in a unique, clinically useful dataset to understand the link between smartphone use, apps and music to mental health, in an important and under-researched cohort. We plan to co-develop interventions based on the findings from this project for future projects to ultimately improve care planning, service delivery, and outcomes.

Project summary

Applications are invited for a fully-funded PhD (fees + stipend) to begin March 2026 on a research project co-developing an AI to predict anxiety and depression in students using smartphone data. The successful candidate will be based at University of Warwick and will be co-supervised by Dr Sagar Jilka and Dr Vivek Furtado.

Project description:

Clinical decisions in mental health are subjective, which limits clinicians’ ability to provide care when anxiety or depressive symptoms start. Smartphones offer unique opportunities to address this, as they unobtrusively provide a continuous stream of unbiased, objective data called ‘digital phenotypes’ (DPs). Over the past decade, smartphone use among young people has increased, concurrently with depressive symptoms in the same age group, with discouraging implications for long-term mental health. AI can use these DPs to characterise complex behaviours to prevent, treat and manage depression.

Depression is highly prevalent, resulting in significant functional impairment, increased risk of suicide and comorbid physical health problems. Mental illness (mainly depression) cost the UK government £117.9 billion, approximately 5% of UK GDP in 2019 through lost productivity.

Previous work has (a) focused on older adults despite young people at greatest risk of depression, (b) been unable to solve data privacy issues, and (c) used invasive and expensive methods. We plan to address these shortcomings to predict mental health risk in university students, an under-researched population, using passively collected DPs.

With your application, please provide a one-page proposed methodology for the PhD, outlining a plan for your PhD research within this area.

The candidate:

The successful applicant will use a mixed-methods approach to the research. They may also conduct relevant systematic and/or meta-analytic reviews to describe the current literature and inform their research. They will undertake qualitative data collection and analysis to understand student perceptions of AI and mental health and acceptability and feasibility of using smartphone data and AI to predict anxiety and depression. They will then use advanced statistical analyses to develop AI models using smartphone data collected during the PhD, focusing on privacy preserving AI such as federated learning. They will undertake work with stakeholders to develop their research and ensure it is meaningful and acceptable to the communities the work benefits. They will communicate with young people and stakeholders to ensure effective dissemination of the work. The successful applicant will join a team of post-doctoral researchers, PhD students, and undergraduate students at the University of Warwick and work closely with relevant networks including the Midlands Mental Health and Neuroscience PhD Programme for Healthcare Professionals.

Required

At least a high 2.1 bachelor’s degree in Neuroscience, Psychology or a related field, or on track to be awarded such a degree before March 1st 2026
At least a high 2.1 grade in a research-based dissertation conducted as part of an undergraduate degree, or equivalent research experience (e.g. through volunteering with a research team)
A strong interest in research and a high level of motivation to develop research ideas
Excellent interpersonal and organisational skills
Ability to code (e.g., in Python, R or similar), visualise and interpret data
Knowledge of machine learning, statistics and a willingness to learn more advanced methods
Some knowledge of qualitative methods and a willingness to learn more
English language proficiency
Ability to work independently when required but to seek supervision appropriately
An understanding of how to work with stakeholder organisations to plan, develop or conduct useful research and demonstration of the skills necessary to do this well

Desirable (but not essential):

A first class bachelor’s degree from a research-intensive University in Neuroscience, Psychology or a related field
A Master’s (with at least Merit) in a relevant field
Knowledge of advanced AI such as deep learning models and privacy preserving AI such as federated learning
Experience of conducting research with individuals from vulnerable populations and/or their families
Experience of working with organisations that hold stakes in the lives of individuals from vulnerable groups, either in a research or other professional context
Ability to communicate complex information clearly
Evidence of being able to disseminate research findings (e.g. via a seminar or conference presentation, or a journal publication)

Application procedure:
Prospective candidates are required to follow University of Warwick guidance. In order for the application to be processed, candidates should submit the research proposal as outlined above, CV, 2 references, and transcript of grades using the online portal: Applying for Postgraduate Study at Warwick.

Further Information:
Please contact Dr Sagar Jilka (sagar.jilka@warwick.ac.uk) for further information and informal enquiries.

Funding notes:

This award, funded by University of Warwick, provides annual funding to cover UK tuition fees, a tax-free stipend and a one-off research training grant of up to £5,000. The funding is provided as contribution to the Midlands Mental Health and Neuroscience PhD Programme for Healthcare Professionals.

The studentship will be awarded on the basis of merit for 3 years of full-time study to commence on 2 March 2026.