Researchers at the Georgia Institute of Technology recently published a report in IEEE Transactions on Biomedical Engineering that demonstrated proof-of-concept of an early-stage home heart health monitoring technology embodied in a bathroom scale. The Georgia Tech engineers employ machine learning to analyze electrocardiograms (ECGs) and ballistocardiograms (BCGs) of patients with heart failure.
The engineers record ECGs and BCGs with a bathroom scale equipped with metal pads within reach of a person standing on the scale. When heart patients step on the scale and touch the pads, the device records both types of data. BCGs measure subtle body movements caused by pulsing blood circulation.
BCGs were used to assess heart health before superior imaging from ECGs was available. Because BCGs are mechanical measurements with multiple interference factors, physicians stopped using ballistocardiograms in favor of easier-to-read ECGs. When the Georgia Tech engineers tested three machine learning algorithms with BCGs, however, they discovered patterns that indicate problems with the heart’s ability to compensate for blood flow irregularities.
Healthy hearts can compensate for reduced blood flow. Physicians describe a patient with an unhealthy heart that cannot manage blood flow irregularities as having “decompensated heart failure.” ECGs do not reveal issues with compensation, but that’s exactly what BCGs are good for.
Humans may not be able to make sense of BCG patterns, but the Georgia Tech researchers found that machine learning algorithms can discriminate BCG variations accurately.
The Georgia Tech engineers’ study showed the mixed technology device’s success with recording and processing data from 43 heart failure patients. The team plans to develop the AI-powered concept further and bring it to market.
At some point in the future, stepping on a bathroom scale may do more than register your weight; if the readings stray from healthy norms, it may message your doctor and alert you to adjust your medication. This sort of timely intervention can reduce the rate of complications that can be far more expensive to treat.
