Machine learning in radiation therapy tested
September 18, 2019
TORONTO – The first-ever patient radiation therapy treatments generated with machine learning in the RayStation treatment planning system (TPS) have been conducted. Patients with localized prostate cancer are being treated with this unique technology at the Princess Margaret Cancer Centre in Toronto, as part of a comprehensive evaluation study.
RaySearch released the machine learning features in RayStation 8B in late December 2018. This ground-breaking technology has been developed by RaySearch’s in-house machine learning department in collaboration with researchers at the Princess Margaret Cancer Centre and Techna Institute, crystalizing years of cutting-edge research led by medical physicist Dr. Tom Purdie and computer scientist Dr. Chris McIntosh.
The features represent the first applications of machine learning in a TPS on the radiation oncology market, producing high quality radiation treatment plans in only minutes, without the need for any user intervention.
Since May 2019, every patient with localized prostate cancer treated at the Princess Margaret has been part of a prospective initiative under the direction of radiation oncologist Dr. Alejandro Berlin.
The initiative was launched after observing excellent clinical results in a retrospective evaluation study conducted during 2018, in which machine learning plans were preferred or deemed equivalent to previous manual plans based on three blinded expert reviewers in 94% of cases.
The ongoing phase of this study presents physicians with two blinded treatment plans: a manually generated plan and a machine learning plan. The selected plan undergoes standard peer-review and quality assurance, and then patients proceed to treatment delivery with the preferred plan.
This world-wide unique endeavor, enabled by RaySearch technology and the team at the Princess Margaret, will provide unique data to quantify the performance and preferability of machine learning plans in the real-world environment.
Dr. Alejandro Berlin (pictured), radiation oncologist at Princess Margaret, says: “It has been really exciting for the team to help materialize this machine learning advancement in the radiation oncology field, including deployment into the clinical realm. Our positive results to date validate our observations about the robustness of this planning solution.”
Johan Löf, founder and CEO, RaySearch, says: “I’m thrilled to hear that patients are being treated with machine learning generated plans from RayStation. These functionalities are the first of its kind and now it is proven that they function just as we anticipated. Our collaboration with Princess Margaret on this project has been fortunate and will lead to better cancer care for more patients.”
About Princess Margaret Cancer Centre
The Princess Margaret is a comprehensive cancer center with over 17000 new patients per year and over 3,000 staff. Its 850,000 square feet house with 202 in-patient beds, over 416,000 square feet of research space and 17 radiation treatment machines on a single site, making it one of the largest comprehensive cancer treatment facilities in the world and the largest radiation treatment center in Canada. Through ongoing research, education and innovation, Princess Margaret continues to be on the frontiers of medical, surgical and radiation oncology, embracing the latest technology and international best-practices and setting standards for patient care.
RaySearch is a medical technology company that develops innovative software solutions to improve cancer care. The company markets the RayStation treatment planning system and RayCare, the next-generation oncology information system, worldwide. Over 2,600 clinics in more than 65 countries use RaySearch’s software to improve life and outcomes for patients. The company was founded in 2000 and the share has been listed on Nasdaq Stockholm since 2003. More information about RaySearch is available at www.raysearchlabs.com.
RayCare is designed to support the complex logistical challenges of modern oncology clinics. It represents the future of oncology information system technology, supporting the vision of one oncology workflow. Many cancer patients receive a combination of treatment types, and RayCare is designed to reflect that. It will efficiently coordinate activities in radiation therapy, chemotherapy and surgery and will offer advanced features for clinical resource optimization, workflow automation and adaptive radiation therapy. RayCare is being developed with tomorrow’s requirements for advanced analytics and decision support in mind.
RayStation is a flexible, innovative treatment planning system, chosen by many of the leading cancer centers worldwide. It combines unique features such as unmatched adaptive therapy capabilities, multi-criteria optimization, market-leading algorithms for IMRT and VMAT optimization with highly accurate dose engines for photon, electron, proton and carbon ion therapy. RayStation supports a wide range of treatment machines, providing one control center for all treatment planning needs and ensuring centers get greater value from existing equipment. RayStation also seamlessly integrates with RayCare, the next-generation oncology information system. By harmonizing the treatment planning, we enable better care for cancer patients worldwide.