Remote monitoring in heart failure: artificial intelligence and the use of remote speech analysis to detect worsening heart failure events

Globally, heart failure (HF) is a leading cause of hospitalization and mortality, primarily among the elderly, and is estimated to affect more than 64 million individuals. Hospitalization for HF represents the largest part of overall medical care expenditures for HF, and hospitalization for HF is associated with high rates of in-hospital and post-discharge morbidity and mortality. Patients discharged from the hospital with a diagnosis of acute decompensated HF have an increased risk for clinical worsening, rehospitalization, and mortality. A major goal for patients with HF is to detect and prevent both first and recurrent hospitalizations. However, detecting and preventing worsening HF events requiring hospitalization and/or pharmacotherapy remains an unmet medical need. Artificial intelligence (AI) is helping us meet this clinical challenge. An example leverages speech processing for the assessment of HF clinical status. In the acute setting, changes in speech measures (SM) can identify the decompensated from the compensated state. A remote monitoring system (HearO™), which includes a mobile speech application (App) to detect worsening HF prior to decompensation events is undergoing evaluation in ambulatory HF patients for reducing the rate of hospitalization. This App is readily downloadable on a smartphone and is user-friendly, and presents an example of how AI-assisted speech signal processing system development may enhance diagnostic accuracy. Preliminary results from clinical trials indicate high rates of sensitivity for detecting HF events along with high rates of adherence. Further elucidation of the effectiveness of this system will be provided by ongoing and planned studies in patients with chronic HF.