Don't miss our weekly PhD newsletter | Sign up now Don't miss our weekly PhD newsletter | Sign up now

  Predictive Diabetes Management: A Self-Learning Human Digital Twin System

   Department of Computer Science

   Friday, June 28, 2024  Funded PhD Project (UK Students Only)

About the Project

This research is a collaborative effort in between Aberystwyth University and Hywel Dda University Health Board (HDUHB).

Project description

The implementation of Human Digital Twin (HDT) within the healthcare system in middle and west Wales represents a transformative step towards enhancing healthcare delivery and patient outcomes across the region. As healthcare systems evolve to meet the demands of modern medicine, the concept of HDT offers a promising framework to revolutionise how patient data is managed, analysed, and utilised within the healthcare ecosystem. By leveraging innovative technologies and data-driven insights, HDT aims to empower healthcare providers, improve clinical decision-making, and ultimately elevate the quality of care delivered to patients throughout middle and west Wales. This project proposes the development of a digital twin-based system for personalized healthcare system in middle and west Wales, representing a pivotal element in the full digital twin development strategy aiming to seamlessly integrate primary and secondary care services.

To initiate the development of this HDT framework, this research project proposes a system specifically designed to detect type-2 diabetes. This focus is particularly relevant as Wales faces a significant public health challenge: nearly 8% of adults over 17 in 2021/22 were diagnosed with diabetes, primarily type-2, representing a concerning 40% increase in cases over the past 12 years. By leveraging patient data from electronic health records, wearables, population health surveys and other potential data streams, the HDT can employ integrative AI techniques such as machine learning, deep learning and data mining techniques to identify individuals at high risk. This allows for early intervention, personalized risk stratification, and predictive analytics to anticipate potential diabetic episodes. This HDT project is designed to continuously improve through periodic recalibration with new patient data. Over time, it will provide increasingly accurate patient predictions and treatment suggestions. To achieve this, the project requires the development of three core modules:

  1. Data Acquisition: This module will gather information from various sources, including imaging, clinical history, demographics, surveys, and data from low-cost wearable devices.
  2. Digital twin modelling: This module will analyze the collected data to build models and simulations that can predict patient outcomes.
  3. Decision Making: This module will leverage the models and simulations to generate treatment suggestions for healthcare professionals.

The learnings and infrastructure developed here will serve as a foundation for the broader HDT system, ultimately personalizing healthcare delivery and improving patient outcomes across middle and west Wales.

Perceived benefits

The implementation of the HDT is expected to yield several significant benefits for the healthcare system in middle and west Wales:

  1. Personalized healthcare: Tailored treatment plans based on individual patient characteristics and data will improve the effectiveness of healthcare interventions.
  2. Real-time monitoring: Continuous monitoring and analysis of patient health will enable early detection of potential health issues and allow for timely intervention.
  3. Proactive healthcare management: By anticipating potential health risks, the HDT can facilitate proactive intervention and prevention measures, reducing the burden of chronic diseases and improving overall health outcomes.
  4. Cost-effectiveness: Improved patient outcomes and reduced healthcare costs are expected to result from the HDT, making it a cost-effective investment in the long run.

Project aims

  1. Establish an innovative foundational HDT that leverages AI techniques to revolutionize healthcare delivery in middle and west Wales.
  2. Implement an initial system focusing on early identification of individuals at high risk of developing type-2 diabetes utilizing data from electronic health records, wearables, population health surveys, and potentially other sources.
  3. Employ AI techniques to provide healthcare providers with predictive analytics, risk stratification tools and automated prediction system for personalized interventions and improved clinical decision-making.
  4. Enable early detection and preventative measures for type-2 diabetes, ultimately aiming to elevate the quality of care delivered to patients throughout the region.

Required skills

  1. Familiarity with relevant programming languages (Python, R) and data science libraries (Scikit-learn, TensorFlow) is essential.
  2. Given the collaborative nature of the project between academia and healthcare institutions, the candidate should have excellent communication and collaboration skills. They should be able to effectively collaborate with researchers, clinicians, and other stakeholders to integrate diverse perspectives and expertise into the research process.
  3. Experience working with databases or big data platforms is a plus.
  4. A background in AI, data wrangling, and model evaluation is a plus.


  • Dr. Praboda Rajapaksha: Lecturer in Health Data Science, Department of Computer Science, Aberystwyth University, and Data Scientist, Hywel Dda University Health Board
  • Mr. Gareth Jenkins: Head of Data Science, Hywel Dda University Health Board

Application deadline:

28th June 2024

Expected start date:

September 2024

How to apply:

For more details about the project, contact the supervisor, Dr. Praboda Rajapaksha ().

To apply for the role, applicants should submit a personal statement, CV, and other supporting documents to Prof Tossapon Boongoen ().

Computer Science (8)

Funding Notes

Aberystwyth University Robertson scholarship

Register your interest for this project

Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.