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

Fully Funded PhD Opportunity: AI for Early Detection of Mood Problems in University Students

Applications are invited for a fully funded PhD (fees and stipend) to begin March 2026 at the University of Warwick. This exciting project, co-supervised by Dr. Sagar Jilka and Dr. Vivek Furtado, focuses on developing a privacy-preserving Artificial Intelligence (AI) tool to predict the worsening of anxiety and depressive symptoms in young people in higher education (YPHE) using ubiquitous smartphone data.

Project Description:

Mental health challenges, particularly anxiety and depression, are highly prevalent among university students and carry significant personal and societal costs, including substantial economic burdens on the UK. Current clinical assessments are often subjective and reactive. This PhD aims to revolutionize early intervention by leveraging digital phenotypes (DPs) - unbiased, continuous data streams from daily smartphone use, social media, and music listening.

Despite the promise of DPs in predicting symptom exacerbation, previous research has typically neglected the highest-risk group, young people, and struggled to overcome critical issues like data privacy and the use of invasive, expensive methods. This project directly addresses these shortcomings. It will utilise readily available, non-invasive DPs to predict mental health risk in an under-researched, yet crucial, university student population. The ultimate goal is to create a clinically useful dataset and co-develop interventions to improve care planning and outcomes.

Research Approach and Methods:

The PhD will employ a mixed-methods approach across three key work packages. Initially, the candidate will qualitatively explore students' understanding of mental health, AI, and DPs, working closely with a student advisory group to ensure the tool is acceptable and meaningful (Work Packages 1 & 2).

The core of the research (Work Package 3) involves a longitudinal study to collect smartphone data and test the validity of the co-developed AI. The data analysis will focus on building sophisticated Deep Learning models, e.g., Long Short-Term Memory (LSTM) networks, to accurately model DPs over time and predict mood deterioration. The project will implement a Federated Learning approach to deploy the AI, ensuring robust privacy preservation of sensitive student data. The successful applicant will undertake advanced statistical analysis and stakeholder engagement to ensure effective and ethical deployment.

The Candidate:

We are seeking a highly motivated candidate with at least a high 2.1 bachelor’s degree (or equivalent research experience) in Neuroscience, Psychology, or a related field, to be awarded before March 1st, 2026. Essential skills include an ability to code (e.g., Python, R) and interpret data, knowledge of machine learning and statistics, and a willingness to learn advanced methods, including qualitative techniques. A strong interest in research, excellent interpersonal skills, and an understanding of stakeholder engagement are vital. Desirable qualities include a Master's degree, experience with advanced AI like deep learning or federated learning, and prior experience conducting research with vulnerable populations.

To apply, candidates must submit a one-page proposed methodology for the PhD, a CV, two references, and a transcript of grades via the University of Warwick online portal. Informal enquiries are welcomed by Dr. Sagar Jilka .

Funding

The award will cover the UK tuition fee level for 4 years, plus a tax-free stipend, currently £20,780 (2025/2026), paid at the prevailing UKRI rate for 3 years of full-time study and a one-off research training grant of up to £5,000.

Advert information