Machine Learning Improves Early Accurate Detection of Lymphedema
Using real-time symptom reports, a well-trained classification algorithm can detect lymphedema more accurately than current clinical methods.
Lymphedema, an adverse effect of breast cancer treatment that causes swelling in the arms or legs, occurs in more than 41% of patients within 10 years of their surgery.
According to a new study published in mHealth, researchers have developed an efficient way to detect lymphedema with machine learning, a type of artificial intelligence.
Lymphedema can be a debilitating adverse effect of breast cancer treatment that can progress to a severe and chronic condition if left untreated. There is no cure for lymphedema, but early detection and treatment can help reduce symptoms and keep it from worsening.
“Clinicians often detect or diagnose lymphedema based on their observation of swelling,” study author Mei R Fu, PhD, RN, FAAN, associate professor of nursing at NYU Rory Meyers College of Nursing, said in a press release. “However, by the time swelling can be observed or measured, lymphedema has typically occurred for some time, which may lead to poor clinical outcomes.”
In the study, the researchers used a web-based tool to collect information from 355 women who had undergone treatment for breast cancer, including surgery. Participants were asked to give demographic and clinical information, including whether they had been diagnosed with lymphedema and whether they were currently experiencing 26 different symptoms.
The researchers performed a standard statistical approach and 5 machine learning approaches to detect lymphedema. According to the findings, all of the machine learning approaches outperformed the statistical procedure. The artificial neural network achieved the best performance, with 93.75% accuracy in classifying patients to have true lymphedema cases or non-lymphedema cases based on the symptoms reported.
According to Dr Fu, this detection accuracy is significantly higher than that achievable by current clinical methods.
Because lymphedema can occur immediately after surgery or up to 20 years later, real-time detection of lymphedema is crucial to achieving timely detection and treatment to prevent the risk of progression to chronic stages.
Additionally, conducting real-time lymphedema assessments can facilitate patients to monitor their status without the burden of unnecessary clinical visits. By assessing their symptoms, the system could alert patients who are at risk to schedule in-person appointments with a health care professional.
“This has the potential to reduce health care costs and optimize the use of health care resources through early lymphedema detection and intervention, which could reduce the risk of lymphedema progressing to more severe stages,” Dr Fu said in the press release.
Overall, the researchers concluded that the use of a well-trained classification algorithm to detect lymphedema based on symptom features is a promising tool that may improve outcomes.
Fu MR, Wang Y, Li C, et al. Machine learning for detection of lymphedema among breast cancer survivors. mHealth. 2018. doi: 10.21037/mhealth.2018.04.02
Machine Learning Helps Detect Lymphedema Among Breast Cancer Survivors [news release]. NYU’s website. https://www.nyu.edu/about/news-publications/news/2018/june/machine-learning-helps-detect-lymphedema-among-breast-cancer-sur.html. Accessed June 11, 2018.