Background:
Advances in wearable and smartphone technologies have opened new possibilities for continuous, real-world monitoring of health and behaviour. While numerous studies have explored digital biomarkers within specific conditions such as depression or epilepsy, there remains a critical gap in understanding shared and disorder-specific digital signatures across mental health and neurological disorders. This PhD project utilises rich multimodal datasets collected across multiple clinical conditions, including Attention-Deficit/Hyperactivity Disorder (ADHD), Eating Disorders, Depression, Multiple Sclerosis (MS), and Epilepsy. These comprehensive datasets offer an exceptional opportunity to investigate behavioural, cognitive, and physiological patterns across diverse patient populations.
Novelty and Importance:
This research moves beyond disorder-specific analyses to develop a cross-condition digital biomarker framework, integrating data from wearables and smartphones to identify common behavioural disruptions (e.g., circadian irregularity, mobility reduction) alongside condition-specific markers (e.g., pre-ictal changes in epilepsy, fatigue in MS, impulsivity in ADHD). By combining data-driven modelling, longitudinal analysis, and machine learning, the project will establish a unified digital health architecture capable of early detection of symptom relapse or episodic events across multiple conditions. The work has strong translational relevance for remote patient monitoring, digital clinical trials, and healthcare policy, contributing toward the standardisation of digital endpoints in real-world healthcare.
Aims and Objectives:
The PhD aims to design and validate a modular, scalable framework for cross-condition digital biomarkers. Specific objectives include:
- Characterising shared and condition-specific behavioural and physiological signatures.
- Developing multimodal models to predict relapse, fatigue, or seizure events.
- Implementing personalised monitoring approaches using individual baselines.
- Validating digital biomarkers against clinical outcomes and informing policy on digital health data standardisation.