Dal Alert!

Receive alerts from Dalhousie by text message.


Lecture: A Mixed-Integer Linear Programming Optimization Model for Capturing Expert Treatment Planning Style in Low-Dose Rate Prostate Brachytherapy

Low dose rate (LDR) brachytherapy is a minimally invasive form of radiation therapy, used to treat prostate cancer, and it involves the permanent implantation of radioactive sources (seeds) inside of the prostate gland. Treatment planning in brachytherapy consists of a decision making process for the placement of radioactive sources in order to deliver an effective dose of radiation to cancerous tissue in the prostate while sparing the surrounding healthy tissue (especially the urethra and rectum). While treatment planning is usually carried out manually by expert planners in the majority of cancer clinics worldwide, such a decision making process can also be automated by modelling it as a mixed-integer linear programming (MILP) problem.

Even though there are several research-based and commercial optimization approaches available today for clinical use, many cancer centres find these to be too slow, inconsistent or unsuitable to integrate into their brachytherapy procedures. In order to fill this existing gap and address the shortcomings of such optimization approaches, we introduce a novel MILP optimization model for interstitial low-dose rate prostate brachytherapy that attempts to mimic the qualities of treatment plans produced manually by expert planners. Our approach involves incorporating a unique set of clinically important constraints, called spatial constraints, that enable us to capture the treatment planning style present at a cancer centre. Furthermore, unlike previous optimization studies, we also attempt to capture the essential aspects of the manual-planning method in order to develop an intuitive optimization approach that expert human planners will find seamless to adopt within their dai! ly practice.

Preliminary results, obtained from a data set involving twenty patients previously treated at the Cross Cancer Institute in Edmonton, show that treatment plans produced through our optimization approach largely capture the qualities and characteristics of manual plans created by expert planners. A highlighting feature of our results is the ability to produce treatment plans in as little as half a minute, which is a noteworthy improvement over the currently employed manual methods that usually take about one to four hours by expert planners. 


Lectures, Seminars




MA 310, 5269 Morris Street 



Additional Information

Speaker: Ege Babadagli, BSc, PhD (Currently a Medical Student at Dalhousie)