Models for Radiation Therapy Patient Scheduling
Sara Frimodig, Christian Schulte.
In Europe, around half of all patients diagnosed with cancer are treated with radiation therapy. To reduce waiting times, optimizing the use of linear accelerators for treatment is crucial. This paper introduces an Integer Programming (IP) and two Constraint Programming (CP) models for the non-block radiotherapy patient scheduling problem. Patients are scheduled considering priority, pattern, duration, and start day of their treatment. The models include expected future patient arrivals. Treatment time of the day is included in the models as time windows which enable more realistic objectives and constraints. The models are thoroughly evaluated for multiple different scenarios, altering: planning day, machine availability, arrival rates, patient backlog, and the number of time windows in a day. The results demonstrate that the CP models find feasible solutions earlier, while the IP model reaches optimality considerably faster.
In: Thomas Schiex, Simon de Givry, editors, Twentyfifth International Conference on Principles and Practice of Constraint Programming, Stamford, CT, USA, volume 11802 of Lecture Notes in Computer Science, pages 421-437. Springer-Verlag, September, 2019. DOI 10.1007/978-3-030-30048-7_25.
Copyright Springer-Verlag, the original publication is available at www.springerlink.com