Using AI for Reducing Surgical Waiting Times

Implementing AI and Optimisation for Effective Scheduling of Elective Surgeries

The project has contributed to the management of RTT patient waiting times at East Suffolk and North Essex NHS Foundation Trust (ESNEFT) by developing machine learning driven solutions to identify how different surgical procedures, patient and clinical resource characterises impact surgery times. These insights were then used to generate simulated theatre schedules which were optimised used computational and mathematical optimisation methods to maximize the utilisation of available resources while considering resource and patient-related constraints.

The project outcomes have demonstrated the potential to assist clinicians and planners in creating optimal schedules that enhance theatre efficiency to enable reduction in RTT patient waiting times.

Project milestones

1. Theatre procedure time predictive model that utilises advanced machine learning techniques together with past surgical procedure-related data from ESNEFT to estimate the surgical procedure time.
2. Simulation model to assess the real-world variability of procedure timing during patient flow through different stages in the pathway, to generate accurate schedules.
3. Optimisation model using Metaheuristic and mathematical optimisation techniques to create optimal schedules for patients in the waiting list.

Partners and funding

This project was run in collaboration and funded by East Suffolk and North Essex NHS Foundation Trust (ESNEFT). The project was awarded "Best Partnership Public or Third Sector" at the Essex KTP Awards 2024.

Publications

• M. S. Sapkota, F. Doctor, H. Herrera, M. Kampouridis and X. Yang, "Machine Learning to Predict Surgery Duration: Towards Implementing AI and Digital Twin for Effective Scheduling," 2024 IEEE International Conference on Medical Artificial Intelligence (MedAI), Chongqing, China, 2024, pp. 435-440