Researchers utilized five different machine learning approaches to accurately spot lymphedema—a negative side effect of breast cancer treatment—which may help detect it earlier and improve treatment.
The study was published in the May edition of the journal mHealth.
"Using a well-trained classification algorithm to detect lymphedema based on real-time symptom reports is a highly promising tool that may improve lymphedema outcomes," said lead author Mei R Fu, PhD, and associate professor of nursing at New York University in an NYU release.
A web-based tool collected information from 355 women who had undergone treatment for breast cancer, including surgery. Participants shared demographic data and clinical information and were asked if they were experiencing any 26 different lymphedema-related symptoms.
Five different machine learning classification algorithms were applied to the data, which were compared to a traditional statistical approach that determined the optimal threshold for the symptom count based on receiver operating curve.
Results showed all five machine learning techniques outperformed the standard approach, with an artificial neural network method achieving 93.75 percent accuracy—outperforming the other four approaches.
Fu and colleagues suggested performing such real-time lymphedema assessments may encourage patients to monitor their condition remotely, which could ultimately eliminate the need for in-person visits and lower increasing healthcare costs.
"This has the potential to reduce healthcare costs and optimize the use of healthcare resources through early lymphedema detection and intervention, which could reduce the risk of lymphedema progressing to more severe stages," Fu said.