Detection of Hypokalemia, Hyponatremia and Hyperkalemia in Heart Failure Patients Using Artificial Intelligence Techniques via ECG

OBJECTIVE
Detection and monitoring of electrolyte imbalance is essential for appropriate treatment of many metabolic diseases. However, there is not yet a tool that can detect such imbalances reliably and noninvasively. Electrolyte disorders, especially in heart failure patients, can cause life-threatening situations, which may develop due to the medications used in the routine treatment of the disease.

METHODS
In this study, we developed a deep learning model (DLM) using electrocardiography (ECG) to detect electrolyte imbalance in heart failure patients and tested its performance in a multicenter study. 17 different centers were included in this study. Heart failure (EF ≤45%) patients whose blood electrolyte values were taken and ECG were taken on the same day were included in the study. The patients were divided into 4 groups: those with normal electrolyte values, those with hypokalemia, those with hyperkalemia and those with hyponatremia. Patients who developed electrolyte disorders due to medications used in heart failure were included in the relevant group. Confidence intervals(CI): We computed 95% CIs for AUROC via stratified bootstrap (2,000 resamples at the patient level), and 95% CIs for Accuracy using the Wilson score interval for binomial proportions.

RESULTS
Accuracy rates of DLM in hyponatremia, hypokalemia and hyperkalemia groups were determined as 83.33%, 95.33% and 95.77%, respectively.

CONCLUSION
The proposed DLM demonstrated high performance in detecting electrolyte imbalance. These results suggest that a DLM can be used for detecting and monitoring electrolyte imbalance using ECG a daily basis.