Resumen: End-stage renal disease (ESRD) patients undergoing hemodialysis (HD) are at high risk of arrhythmias and sudden cardiac death as a result of blood potassium concentration ([K+ ]) changes. The aim of this study is to investigate if dw, a time-warping-based electrocardiogram (ECG) biomarker of T-wave morphology changes, reflects [K+] evolution in HD patients, facilitating noninvasive [K+] monitoring and avoiding in-hospital blood tests analysis. 48-hour ECGs and a set of hourly-collected blood samples from 12 ESRD patients were acquired and analyzed. dw was calculated between a reference T-wave, measured at the end of the HD session, and the T-waves corresponding to each hour along the whole HD session, when [K+] was measured from blood samples. The values of dw correlated with the relative variations in [K+] with respect to the reference value (end of HD, ¿[K+ ]), with a median (interquartile) correlation coefficient of 0.90 (0.30), evidencing a strong relation between them. Our findings support the use of dw as a surrogate of ¿[K+], suggesting a potential use of dw for non-invasive hyperkalemia monitoring both in hospital and ambulatory settings. Idioma: Inglés DOI: 10.23919/CinC49843.2019.9005904 Año: 2019 Publicado en: Computing in Cardiology 46 (2019), [4 pp] ISSN: 2325-8861 Factor impacto SCIMAGO: 0.296 - Computer Science (miscellaneous) - Cardiology and Cardiovascular Medicine