– A medical monitoring scale uses machine learning to track symptoms in heart failure patients, and could potentially lead to fewer hospitalizations and lower care costs, according to research published in IEEE Transactions on Biomedical Engineering.
The experimental device records an electrocardiogram from a patient’s fingers, as well as circulation pulsing throughout the body, known as a ballistocardiogram (BCG). The device then uses machine learning tools to determine whether the patient’s heart failure symptoms have worsened.
Researchers at Georgia Institute of Technology developed the scale and reported proof-of-concept success in recording and processing data from 43 patients with heart failure. A future version of the device would ideally notify a doctor, who would tell the patient to adjust his or her medication accordingly.
The new device could help improve care for the more than 6.5 million Americans suffering from heart failure, who often have to endure multiple hospitalizations because of the slow, progressive nature of the disease.
Home monitoring can reduce hospitalizations, but currently requires an invasive procedure. This new device would be much simpler to use and would cost much less.
The scale performed well not only in hospital tests, but also in in-home tests, showing that the solution is a viable alternative to other methods of heart failure monitoring.
“Our work is the first time that BCGs have been used to classify the status of heart failure patients,” said Omer Inan, the study’s principal investigator and an associate professor in Georgia Tech’s School of Electrical and Computer Engineering.
Researchers believe the key to the scale’s success lies in the measuring of both EKG and BCG waves. While the EKG signal is electrical, the BCG is a mechanical signal, and faces a lot of interference in the body from tissue variations and muscle movement.
“The ECG (EKG) has characteristic waves that clinicians have understood for 100 years, and now, computers read it a lot of the time,” Inan said. “Elements of the BCG signal aren’t really known well yet, and they haven’t been measured in patients with heart failure very much at all.”
Researchers used three machine learning algorithms to process BCG signals, showing patterns that differ when a patient’s heart failure is healthier from when it is decompensated, or unhealthier.
“In someone with decompensated heart failure, the cardiovascular system can no longer compensate for the reduced heart function, and then the flow of blood through the arteries is more disorderly, and we see it in the mechanical signal of the BCG. That difference does not show up in the ECG because it’s an electrical signal,” Inan said.
“The most important characteristic was the degree to which the BCG is variable, which would mean inconsistent blood flow. If you chop up the recording into 20-second intervals and the individual segments differ from each other a lot, that’s a good marker of decompensation.”
Machine learning has played a big role in heart failure and cardiovascular research recently. A team from Brigham and Women’s Hospital and UT Southwestern have developed a machine learning algorithm that can accurately predict heart failure among patients with diabetes.
“We hope that this risk score can be useful to clinicians on the ground – primary care physicians, endocrinologists, nephrologists, and cardiologists – who are caring for patients with diabetes and thinking about what strategies can be used to help them,” said the study’s co-first author Muthiah Vaduganathan, MD, MPH, a cardiologist at Brigham and Women’s Hospital.
In a separate study, researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) built a machine learning tool that can estimate a patient’s risk of cardiovascular death by measuring the electric activity of their heart.
“Machine learning is particularly good at identifying patterns, which is deeply relevant to assessing patient risk,” said Divya Shanmugam, lead author on the paper about the model. “Risk scores are useful for communicating patient state, which is valuable in making efficient care decisions.”
Credit: Google News