Enhancing Healthcare Access using Federated Learning
Research type
Research Study
Full title
Enhancing Healthcare Access using Federated Learning
IRAS ID
332557
Contact name
Aaisha Makkar
Contact email
Sponsor organisation
University of Derby
Clinicaltrials.gov Identifier
N/A, N/A
Duration of Study in the UK
4 years, 11 months, 30 days
Research summary
Accessibility to appointments at general practitioner (GP) surgeries are at crisis point since the COVID-19 pandemic, with patients reporting increased waiting times, difficulty in booking appointments, and challenges in accessing the right services on time. To address these issues and enhance healthcare accessibility, this research project focuses on Federated Learning applications in the healthcare sector, particularly concerning common real-world problems related to GP visits. A review of existing appointment scheduling systems in several GP clinics highlighted problems, including the lack of appropriate scheduling mechanisms and dependence on obsolete methodologies such as “the lottery system” or “first come first served”. We aim to develop a research-based model for GP appointment scheduling recognizing the necessity for a faster and more efficient approach, especially given the current state of technology. The model will be designed to be privacy-preserving, user-friendly, and seamlessly integrated with existing GP appointment booking systems. To this end, we will be using the concept of Federated Learning to develop a digital tool in collaboration with West Park Surgery, utilizing its data for model development. This approach ensures that sensitive medical information is not shared and only the trained parameters of the model (not the actual data) will be used to optimize the model. This approach guarantees that patient data remains protected while improving the accuracy of the application. The local model will use data from the patient's medical history, test results, and other relevant information to calculate the severity score. This score will be used to determine the priority of the patient's appointment. Medical information and the existing appointment system will be informed through our collaborations with our GP partner.
Our overall goal is to improve patient care and reduce the burden on GP services. By prioritizing appointments reflected in the scoring system, we aim to ensure that patients with the most urgent health concerns receive the attention they need as quickly as possible.REC name
South Central - Berkshire B Research Ethics Committee
REC reference
24/SC/0144
Date of REC Opinion
20 Jun 2024
REC opinion
Further Information Favourable Opinion