Heart failure (HF) is a complex disease with high-mortality rates, affecting 64.3 million people worldwide, with a median prevalence of 1.72% in Europe and 3.9% in Lithuania. Electrolyte imbalance is a common pro-arrhythmogenic complication of HF with the potential to instigate sudden cardiac death if not promptly corrected. Unfortunately, electrolyte imbalance is usually silent and virtually undetectable without a blood test until electrolytes reach dangerous levels, precluding patients from seeking preemptive medical treatment before the onset of life-threatening events. Regular blood testing could avert unfavorable outcomes but is expensive and unfeasible outside clinical settings. Thus, inexpensive and ideally non-invasive technologies for ambulatory blood electrolyte monitoring would be of clinical importance. Recent studies demonstrated the potential of electrocardiograms (ECGs) as a non-invasive diagnostic tool for electrolyte imbalance in hemodialysis patients, suggesting that ECG-based methods may be viable for blood electrolyte monitoring in HF patients. Despite their promising results, none of these technological solutions were evaluated in HF patients, and nearly all require 12-lead or precordial-lead ECG systems, which are impractical for ambulatory monitoring. Accordingly, the project smartQRST aims to develop ECG-based methods for facilitating non-invasive blood electrolyte monitoring in HF patients. The methods will rely on deep-learning models to derive ECG features like the spatial QRS-T angle from sets of reduced-lead ECGs that wearable devices can register. By enabling reduced-lead estimation of well-known ECG features instead of 12- or precordial-leads, the proposed deep learning methods will allow exploring the clinical value of such features as prospective bloodless markers of electrolyte levels.
KTU Research and Innovation Fund
During the project, it will be developed methods for facilitating non-invasive ambulatory blood electrolyte monitoring in heart failure patients.
Period of project implementation: 2023-04-11 - 2023-12-31
Project partners: Lithuanian University of Health Sciences