Dinamika i mjere entropije srčanog ritma

master's thesis

• Title (english): Heart rate dynamics and entropy measures
• Author: Maria Krajačić
• Mentor: Davor Horvatić (mentor)
• Committee member: Davor Horvatić (predsjednik povjerenstva)
• Committee member: Nikola Poljak (član povjerenstva)
• Committee member: Tomislav Marketin (član povjerenstva)
• Committee member: Mihael Makek (član povjerenstva)
• Committee member: Nenad Pavin (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Physics); Zagreb
• Defense date and country: 2018-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Physics

Abstract

Električni signali proizvedeni u ljudskom organizmu (EKG, EEG, CTG) danas su dio standardne medicinske dijagnostike. No, fizikalni potpis uzroka tih signala ima veći doseg od onog koji se koristi trenutno u standardnoj dijagnostici. Ideja ovog rada je istražiti potencijal varijabli izvedenih iz analize vremenskih serija navedenih fizioloških signala. U radu koristimo analizu fluktuacija kojima je uklonjen trend za određivanje dugodosežnosti korelacija u pojedinom signalu i višeskalnu entropiju uzorka pomoću koje se dobiva informacija o kompleksnosti vremenskih serija. Analiza fluktuacija kojima je uklonjen trend koja pokazuje da kod ispitanika sa zdravim srcem postoje dugodosežne korelacije, a kod ispitanika s kongestivnim zatajenjem srca pokazuje da postoji fenomen prijelaza koji indicira različito ponašanje na malim i velikim skalama. Dodatno je pokazana primjena metode na određivanje uspješnosti terapije. Račun višeskalne entropije uzorka pokazuje da imamo veće vrijednosti te mjere za dinamiku zdravog srca, odnosno da je kompleksnost takvih vremenskih serija veća od kompleksnosti onih koje odgovaraju određenim patološkim stanjima. Osim toga, u analizi dinamike zdravog srca pokazano je da iznos entropije uzorka ovisi o duljini vremenske serije-kraće vremenske serije daju veću vrijednost entropije, a manja vrijednost entropije odgovara većim serijama. Analiza podataka koji odgovaraju asimptomatskoj ventrikularnoj aritmiji i kongestivnom zatajenju srca pokazala je da su pripadne vremenske serije manje kompleksne, a određeni lijekovi koji se koriste za liječenje aritmije smanjuju kompleksnost vremenskih serija.

Abstract (english)

Electrical signals produced in the human body (ECG, EEG, CTG) are now part of standard medical diagnostics. However, the signature, of the dynamics that can be derived from physics, in these signals has a greater capacity than the one currently used in standard diagnostics. The idea of this thesis is to explore the potential of variables derived from the analysis of a time series of said physiological signals. This paper uses a detrended fluctuation analysis that eliminates the trend for determining correlations in each signal and multipath entropy of the cause by which information on the complexity of time series is obtained. Analyzing the fluctuations shows that long-range correlations exist in heart rate of a healthy people. In congestive heart failure, we see a crossover phenomenon that indicates different dynamics for small and large scales. In addition, a fluctuation analysis is applied to therapy evaluation. Multiscale sample entropy method shows that we have greater entropy values of this measure for the healthy heart’s dynamic. The complexity of these signals is greater than the complexity of those corresponding to certain pathological conditions. In addition, in the analysis of heart rate dynamics, it is shown that the value of sample entropy depends on the length of the time series - for shorter time series we have bigger entropy value, and for longer time series value decreases. Data analysis that corresponds to asymptomatic ventricular arrhythmia and congestive heart failure has shown that the associated time series are less complex and certain drugs used to treat arrhythmia reduce the complexity of time series.

• Keywords: EKG; analiza fluktuacija kojima je uklonjen trend; entropija; višeskalna entropija uzorka
• Keywords (english): ECG; detrended fluctuation analysis; entropy; multiscale sample entropy
• Language: croatian
• URN:NBN: urn:nbn:hr:217:385733
• Study programme: Title: Physics; specializations in: Education; Course: Education; Study programme type: university; Study level: integrated undergraduate and graduate; Academic / professional title: magistar/magistra edukacije fizike (magistar/magistra edukacije fizike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2018-07-17 12:44:30