Desarrollo de Algoritmia para Dispositivos de Monitorización Cardíaca de Nueva Generación

Abstract

Background In recent years, the progress experimented by the electrocardiographic and electrophysiologic devices has provided doctors with several tools for diseases diagnosis. Currently, thanks to the progression in the equipment computing capacity and the development of recent modern telemonitoring systems, such as wearables devices, a new world of possibilities has been opened, which has led to interest resurgence in this field. This work is divided into three main areas related to cardiac digital signal processing techniques, namely, development of algorithm for long-term monitoring (LTM) devices, creation of new diagnostic and prognostic risk indexes, and industry collaboration to develop analysis algorithm for a commercial telemedicine device. Traditionally, the way to treat the quality of an electrocardiogram (ECG) record has been the use of filtering stages that improve classical signal quality metrics, such as the signal-to-noise ratio (SNR). The problem of this paradigm is the lack of a medical criterion that allows knowing parts of the registry are valid from a clinical point of view. This problem becomes relevant in LTM records, since the effort required to analyze them, in temporary terms, is high. The knowledge of the signal sections that are clinically analyzable will save time and effort to responsible clinicians of analyzing these records. On the other hand, although the literature is extensive in beat detection algorithms applied to ECG and Holter, in order to process LTM records, an update in all signals analysis phases must be carried out, since they need special treatment due to the huge volume of data. Therefore, within this new scenario it is necessary to update old methods and develop new ones that perform these calculations. Aditionally, Recent researches have proven the relationship between the appearance of fragmented QRS and various diseases such as cardiac sarcoidosis, acute coronary syndrome, arrhythmogenic heart disease, Brugada syndrome, and hypertrophic cardiomyopathy (HCM). In this Thesis, we have focused on HCM because it can present both fibrosis and fragmentation. Both of them are associated with the risk of developing different life-threatening arrhythmias, moreover, they are also related to the aforementioned diseases.