Primary supervisor

Jesmin Nahar

Background

Endometriosis and Polycystic Ovary Syndrome (PCOS) are complex conditions that can involve diverse symptoms and clinical characteristics. Delays in diagnosis and differences in symptom presentation highlight the potential value of data-driven approaches to support earlier identification and risk assessment.

Research Aim

This research aims to investigate Artificial Intelligence (AI) and Machine Learning (ML) methods for the early prediction and risk stratification of endometriosis and PCOS, with the goal of supporting earlier diagnosis and more personalised healthcare.

Methodology

The study will explore clinical, demographic, laboratory and/or other relevant health data to identify patterns associated with disease risk. Machine-learning and, where appropriate, deep-learning models will be developed and compared for prediction and risk classification.

AI Applications

The research will focus on early prediction, risk stratification, explainable AI (XAI), personalised decision support and, where suitable longitudinal data are available, forecasting disease-related outcomes.

Expected Outcome

The project aims to develop an interpretable and reliable AI-enabled framework that can support healthcare professionals in identifying individuals at higher risk and facilitating earlier clinical assessment and personalised healthcare. The system will be designed to support rather than replace clinical diagnosis and professional judgement.

Aim/outline

Aim

To investigate the application of Artificial Intelligence (AI) and Machine Learning (ML) for the early prediction and risk stratification of Polycystic Ovary Syndrome (PCOS), supporting earlier clinical assessment and more personalised healthcare.

Outline

The research will:

  • Analyse relevant clinical, demographic, hormonal, metabolic and/or lifestyle data.
  • Identify patterns and risk factors associated with PCOS using machine-learning techniques.
  • Develop and compare ML models for PCOS prediction and risk classification.
  • Evaluate model performance using appropriate metrics such as sensitivity, specificity, precision, recall, F1-score and ROC-AUC.
  • Apply Explainable AI (XAI) to identify important factors contributing to predictions.
  • Explore AI-based clinical decision support to assist with earlier risk identification.
  • Where suitable longitudinal data are available, investigate forecasting of PCOS-related outcomes and risk trajectories.
  • Consider model robustness, fairness, generalisability and responsible use of AI in healthcare.

Required knowledge

Required Knowledge

Students should have a basic to intermediate understanding of:

  • Artificial Intelligence and Machine Learning
  • Python and/or R programming
  • Data preprocessing, cleaning and feature engineering
  • Supervised machine-learning methods, particularly classification and prediction
  • Basic statistics and data analysis
  • Model evaluation and performance metrics
  • Basic knowledge of healthcare or biomedical data is desirable
  • An interest in AI applications in women's health, explainable AI and personalised healthcare