In a recent study published in Nature Medicine, researchers at the Princess Margaret Cancer Centre offer evidence for the effectiveness of a computer program that can generate effective radiation treatment options for cancer patients.
In a study of 100 patients, researchers had physicians review and evaluate the clinical acceptableness of physician versus computer-generated treatment plans. Evaluators did not know which treatment plans were physician or computer-generated. Results showed that an overwhelming percentage of computer-generated plans were deemed clinically acceptable (89%) and were more often selected as an appropriate treatment plan than physician-generated plans (72%).
Researchers also compared the evaluation and selection of treatment plans in patients who had previously been treated versus for those who had not received any treatment. While 83% of computer-generated plans were selected for patients who had already received treatment, only about 61% were chosen for patients who had not received treatment.
The computer program studied in this study works similarly to other machines utilizing algorithmic learning. The program uses information about similar patient cases to draw conclusions about an effective course of treatment. The use of computer programs for these purposes offers doctors and clinicians potential benefits, including increased efficiency in treatment decision-making.
However, researchers were quick to note that in a real-world setting, the efficacy of computer-generated treatment has a human issue: whether computer-generated treatment plans can be trusted or not.
"Once you put ML[machine learning]-generated treatments in the hands of people who are relying upon it to make real clinical decisions about their patients, that preference towards ML may drop. There can be a disconnect between what's happening in a lab-type of setting and a clinical one," said Dr. Purdie an Associate Professor in the Department of Radiation Oncology at the University of Toronto.
The computer program used in the study is currently being used to generate treatment plans for prostate cancer patients at the Princess Margaret Centre.
Source: Science Daily