Technology, with machine learning algorithms, analyzes cardiac response to submaximal exercise
TUESDAY, Jan. 16, 2018 (HealthDay News) — Wearable technology that records cardiac function, along with machine learning algorithms, can assess compensated and decompensated heart failure states, according to a study published online Jan. 12 in Circulation: Heart Failure.
Omer T. Inan, Ph.D., from the Georgia Institute of Technology in Atlanta, and colleagues assessed a wearable electrocardiogram and seismocardiogram sensing patch among 32 patients with compensated (outpatient) and 13 with decompensated (hospitalized) heart failure. The protocol consisted of patients standing at rest for an initial recording, performing a six-minute walk test, and then standing at rest for five minutes of recovery. This was performed at the time of the outpatient visit or at admission and discharge during a heart failure hospitalization.
The researchers found that a graph similarity score could assess heart failure patient state and correlates to clinical improvement in the 45 patients. There was a significant difference in the graph similarity score metric between decompensated and compensated heart failure (44.4 versus 35.2). For the six decompensated patients with longitudinal data (admission to discharge), there was a significant change in graph similarity score (44 versus 35).
“Wearable technologies recording cardiac function and machine learning algorithms can assess compensated and decompensated heart failure states by analyzing cardiac response to submaximal exercise,” the authors write. “These techniques can be tested in the future to track the clinical status of outpatients with heart failure and their response to pharmacological interventions.”
One author disclosed financial ties to Physiowave Inc.
Copyright © 2018 HealthDay. All rights reserved.
