Detecting heart failure using wearables: a pilot study

Abstract

Objective: Heart failure (HF) can be difficult to diagnose by physical examination alone. We examined whether wristband technologies may facilitate more accurate bedside testing. Approach: We studied on a cohort of 97 monitored in-patients and performed a cross-sectional analysis to predict HF with data from the wearable and other clinically available data. We recorded photoplethysmography (PPG) and accelerometry data using the wearable at 128 samples per second for 5 min. HF diagnosis was ascertained via chart review. We extracted four features of beat-to-beat variability and signal quality, and used them as inputs to a machine learning classification algorithm. Main results: The median [interquartile] age was 60 [51 68] years, 65% were men, and 54% had heart failure; in addition, 30% had acutely decompensated HF. The best 10-fold cross-validated testing performance for the diagnosis of HF was achieved using a support vector machine. The waveform-based features alone achieved a pooled test area under the curve (AUC) of 0.80; when a high-sensitivity cut-point (90%) was chosen, the specificity was 50%. When adding demographics, medical history, and vital signs, the AUC improved to 0.87, and specificity improved to 72% (90% sensitivity). Significance: In a cohort of monitored in-patients, we were able to build an HF classifier from data gathered on a wristband wearable. To our knowledge, this is the first study to demonstrate an algorithm using wristband technology to classify HF patients. This supports the use of such a device as an adjunct tool in bedside diagnostic evaluation and risk stratification.

Machine Learning for Cardiovascular Outcomes From Wearable Data: Systematic Review From a Technology Readiness Level Point of View

“…Raw data could also be used as features upon which a neural network can be used to automatically learn informative features [46]. Next to the features based on the sensed signal, demographic information could be used to provide more context [28,47]. Benchmark studies mostly use raw features (using the same data set) and were, therefore, excluded from this study.…”

“…Less than half (N=13) of the studies were reported with such placements, of which 8 (62%) studies acquired one modality: 3 (23%) studies acquired wrist-based ECGs [18,21,22], 2 (15%) studies acquired wrist-based PPGs [17,23], and 3 (23%) studies acquired finger-based PPGs [24,30,37]. Of the 13 studies, the remaining 5 (39%) studies acquired wrist-based multimodal data: 4 (31%) studies acquired PPGs and accelerometer data [19,20,29,47] and 1 (8%) study acquired both ECGs and PPGs [25]. Thus, the wrist and finger severely limited the additional modalities that were measured (usually only acceleration), although wearables were shown to be able to measure increasing number of modalities [10].…”

Novel Beat-wise Classification Algorithm of ECG to Monitor Progression of Cardiac Health

“…Monitoring for Wearables P. Krishnan, Member, IEEE, V. Rajagopalan, and B.I. Morshed, Sr., Member, IEEE C subjects have less heart rate variability and the PPG sensor reflects oxygen saturation [12]. Heart rate estimation and accuracy of PPG and accelerometer can be improved using Adaptive Neural Network filtering [13].…”

A current and future outlook on upcoming technologies in remote monitoring of patients with heart failure

“…There is currently a heightened focus on wearable monitoring technology both for detecting atrial fibrillation or for preventing hospitalization due to decompensated HF by observing early changes occurring before overt acute HF takes place 57,58 …”