Pre-Emption of Affliction Severity Using HRV Measurements from a Smart Wearable; Case-Study on SARS-Cov-2 Symptoms

Area of Science:

• Biomedical Engineering
• Digital Health
• Wearable Technology

Background:

• Smart wristbands and watches are increasingly used for fitness tracking.
• Their potential in healthcare, particularly for early disease detection, is emerging.
• Continuous monitoring of physiological data via wearables offers new diagnostic avenues.

Purpose of the Study:

• To investigate the use of heart rate variability (HRV) from smartwatches to predict illness onset.
• To develop and validate an algorithm for preempting the worsening of afflictions using HRV.
• To apply this methodology to SARS-CoV-2 as a case study.

Main Methods:

• Utilized data from a Welltory study involving smartwatches (Apple Watch, Fitbit, Garmin) and SARS-CoV-2 symptoms.
• Trained a Hidden Markov Model (HMM) using HRV components (NN intervals, rMSSD, LF, HF, VLF) as internal states.
• Employed the Viterbi algorithm to determine probable sequences of hidden states.

Main Results:

• The HMM confirmed that a consistent decline in HRV components indicated a higher probability of symptom onset or worsening.
• Emission probabilities demonstrated a significant correlation between specific HRV patterns and illness progression.
• Viterbi algorithm confirmed the predictive capability of the model.

Conclusions:

• HRV analysis via smart wearables can potentially preempt the onset or worsening of diseases like SARS-CoV-2.
• Early detection through such algorithms can reduce complications, mortality, and disease spread.
• This approach supports intelligence-backed decisions for proactive public health management.