Researchers out of the University of Surrey in the U.K. created an artificial intelligence (AI) platform that can predict which cancer patients are most at risk for experiencing common symptoms associated with the disease.
In the study, published Dec. 31 in PLOS One, the team collected data from two different time points from 799 oncology cases taken over the course of chemotherapy. Two machine learning algorithms were compared on their ability to predict the severity of sleep disturbance, anxiety and depression compared to real value scores.
Overall, first author Nikolaos Papachristou, with Surrey, and colleagues found the differences between the predicted and real values were “not meaningful,” and that both methods “produced equivalent results for all three symptoms."
The study may be important in that depression occurs in up to 60 percent of cancer patients, the authors wrote. Similarly, up to 53 percent of patients report feelings of anxiety during treatment and nearly half of patients experience both symptoms.
Papachristou and colleagues acknowledged their study was exploratory, but believe their research can be used to develop computational tools which may help clinicians identify risk profiles for patients. This knowledge may allow for customized treatment plants and a better quality of life.
“These exciting results show that there is an opportunity for machine learning techniques to make a real difference in the lives of people living with cancer,” said contributing author Payam Barnaghi, professor of machine intelligence at Surrey, in a news release.