Several mechanisms are responsible for cardiorespiratory interactions observed in humans. The action of these mechanisms results in specific patterns in heart rate variability (HRV) and affects the interaction between heart and respiratory activities. The four main types of phenomena resulting from the interactions between heart and respiratory system are: i) respiratory sinus arrhythmia (RSA); ii) cardioventilatory coupling; iii) cardiorespiratory phase synchronization; iv) cardiorespiratory frequency synchronization. The aim of this thesis is to describe and quantify different aspects of cardiorespiratory interactions employing a variety of methods from literature, adapted and optimized for the usual experimental settings in which HRV and respiratory signal are commonly acquired. Six analytical methods were exploited for this purpose assessing transfer entropy (TE), cross-conditional entropy via normalized corrected cross-conditional entropy (NCCCE), squared coherence (K2), cardioventilatory coupling via normalized Shannon entropy (NSE) of the time interval between QRS complex and inspiratory, or expiratory, onsets, phase synchronization via a synchronization index (SYNC%) and pulse-respiration quotient (PRQ). These approaches were employed with the goal of testing the effects of a sympathetic challenge, namely postural stimuli like head-up tilt (TILT) and active standing (STAND), on cardiorespiratory interactions. The proposed approaches were tested on three protocols: i) amateur athletes undergoing an inspiratory muscle training (IMT) during supine rest (REST) and STAND; ii) healthy volunteers undergoing a prolonged bed rest deconditioning (HDBR), during REST and TILT; iii) patients suffering from postural orthostatic tachycardia syndrome (POTS), during REST and TILT, at baseline and at one-year follow-up. The most important findings of the present doctoral thesis concern the effect of postural stimuli on cardiorespiratory interactions in health and disease. Indeed, all proposed indexes gave a coherent view of cardiorespiratory interaction strength in response to the orthostatic challenge, as it decreased in all protocols. However, the statistical power of the indexes was different. TE and K2 appeared to be particularly weak in detecting the effect of postural challenge on cardiorespiratory interactions. NCCCE, NSE and SYNC% exhibited much stronger ability in this regard, while PRQ seemed too closely related to heart rate, in presence of no significant modification of the respiratory rate. Conversely, all indexes appeared to be weak in detecting the chronic effects of IMT and HDBR on a healthy population and the long-term consequences of the clinical management in POTS patients. The thesis concludes that the different aspects of cardiorespiratory interactions can be modified acutely but the chronic effects of a long-term treatment or intervention on the magnitude of cardiorespiratory interactions are negligible and/or could be confused with the variability of markers. Considerations about the methodological dissimilarities and differences in effectiveness of the proposed indexes suggest that the simultaneous exploitation of all bivariate methodologies in cardiorespiratory studies is advantageous, as different aspects of cardiorespiratory interactions can be evaluated concurrently. This simultaneous evaluation can be carried out with a relatively negligible computational cost and in applicative contexts when only an ECG signal is available.
Molteplici meccanismi sono responsabili delle interazioni cardiorespiratorie osservate nell’uomo. L’azione di questi meccanismi risulta in specifici ritmi di variabilità cardiaca (HRV) e ha effetti sulle interazioni tra attività cardiaca e respiratoria. I quattro principali tipi di fenomeni che derivano dalle interazioni tra cuore e sistema respiratorio sono: i) aritmia respiratoria sinusale (RSA); ii) accoppiamento cardioventilatorio; iii) sincronizzazione cardiorespiratoria in fase; iv) sincronizzazione cardiorespiratoria in frequenza. L’obiettivo di questa tesi è la descrizione e quantificazione di diversi aspetti delle interazioni cardiorespiratorie tramite l’utilizzo di una varietà di metodologie derivate dalla letteratura, adattate e ottimizzate per i tipici contesti sperimentali in cui HRV e segnale respiratorio sono comunemente acquisiti. Sei metodi analitici sono stati sfruttati a questo scopo per valutare l’entropia di trasferimento (TE), l’entropia cross-condizionata tramite entropia corretta cross-condizionata normalizzata (NCCCE), la coerenza quadratica (K2), l’accoppiamento cardioventilatorio tramite entropia normalizzata di Shannon (NSE) dell’intervallo temporale tra complesso QRS e inizio di fase inspiratoria o espiratoria, la sincronizzazione in fase tramite un indice di sincronizzazione (SYNC%) e il quoziente pulsazione-respirazione (PRQ). Questi approcci sono stati utilizzati con la finalità di testare gli effetti di uno stimolo simpatico, ovvero stimoli posturali quali l’head-up tilt (TILT) e l’ortostatismo attivo (STAND) sulle interazioni cardiorespiratorie. Gli approcci proposti sono stati testati in tre protocolli: i) atleti amatoriali sottoposti a un allenamento muscolare inspiratorio (IMT) durante clinostatismo supino (REST) e STAND; ii) volontari sani sottoposti a un decondizionamento da allettamento prolungato (HDBR), durante REST e TILT; iii) pazienti affetti da sindrome da tachicardia posturale ortostatica (POTS), durante REST e TILT, in condizione basale e durante approfondimento un anno dopo. I risultati principali della presente tesi di dottorato concernono l’effetto degli stimoli posturali sulle interazioni cardiorespiratorie in soggetti sani e patologici. Infatti, tutti gli indici proposti danno una visione coerente dell’intensità dell’interazione cardiorespiratoria in risposta a uno stimolo ortostatico, in quanto essa diminuisce in tutti i protocolli. Tuttavia, il potere statistico degli indici è differente. TE e K2 appaiono essere particolarmente deboli nell’identificare l’effetto dello stimolo posturale sulle interazioni cardiorespiratorie. NCCCE, NSE, e SYNC% dimostrano una capacità molto maggiore a tale riguardo, mentre PRQ appare troppo intimamente collegata alla frequenza cardiaca, in assenza di cambiamenti significativi della frequenza respiratoria. Per contro, tutti gli indici appaiono deboli nell’identificare gli effetti cronici di IMT e HDBR in una popolazione sana o le conseguenze croniche della gestione clinica in pazienti POTS. La tesi conclude che diversi aspetti delle interazioni cardiorespiratorie possono essere modificati in modo acuto ma gli effetti cronici di un trattamento o intervento a lungo termine sono irrisori sulla magnitudine delle interazioni cardiorespiratorie e/o possono essere confusi con la variabilità intrinseca degli indici. Considerazioni sulle differenze metodologiche e sull’efficacia degli indici proposti suggeriscono che un utilizzo simultaneo di molteplici metodi bivariati è vantaggiosa negli studi cardiorespiratori, in quanto diversi aspetti delle interazioni cardiorespiratorie possono essere valutati contemporaneamente. Questa valutazione simultanea può essere effettuata a un costo computazionale trascurabile e in contesti applicativi in cui il solo segnale ECG è disponibile.
ESTIMATING CARDIORESPIRATORY COUPLING FROM SPONTANEOUS VARIABILITY IN HEALTH AND PATHOLOGY
CAIRO, BEATRICE
2021
Abstract
Several mechanisms are responsible for cardiorespiratory interactions observed in humans. The action of these mechanisms results in specific patterns in heart rate variability (HRV) and affects the interaction between heart and respiratory activities. The four main types of phenomena resulting from the interactions between heart and respiratory system are: i) respiratory sinus arrhythmia (RSA); ii) cardioventilatory coupling; iii) cardiorespiratory phase synchronization; iv) cardiorespiratory frequency synchronization. The aim of this thesis is to describe and quantify different aspects of cardiorespiratory interactions employing a variety of methods from literature, adapted and optimized for the usual experimental settings in which HRV and respiratory signal are commonly acquired. Six analytical methods were exploited for this purpose assessing transfer entropy (TE), cross-conditional entropy via normalized corrected cross-conditional entropy (NCCCE), squared coherence (K2), cardioventilatory coupling via normalized Shannon entropy (NSE) of the time interval between QRS complex and inspiratory, or expiratory, onsets, phase synchronization via a synchronization index (SYNC%) and pulse-respiration quotient (PRQ). These approaches were employed with the goal of testing the effects of a sympathetic challenge, namely postural stimuli like head-up tilt (TILT) and active standing (STAND), on cardiorespiratory interactions. The proposed approaches were tested on three protocols: i) amateur athletes undergoing an inspiratory muscle training (IMT) during supine rest (REST) and STAND; ii) healthy volunteers undergoing a prolonged bed rest deconditioning (HDBR), during REST and TILT; iii) patients suffering from postural orthostatic tachycardia syndrome (POTS), during REST and TILT, at baseline and at one-year follow-up. The most important findings of the present doctoral thesis concern the effect of postural stimuli on cardiorespiratory interactions in health and disease. Indeed, all proposed indexes gave a coherent view of cardiorespiratory interaction strength in response to the orthostatic challenge, as it decreased in all protocols. However, the statistical power of the indexes was different. TE and K2 appeared to be particularly weak in detecting the effect of postural challenge on cardiorespiratory interactions. NCCCE, NSE and SYNC% exhibited much stronger ability in this regard, while PRQ seemed too closely related to heart rate, in presence of no significant modification of the respiratory rate. Conversely, all indexes appeared to be weak in detecting the chronic effects of IMT and HDBR on a healthy population and the long-term consequences of the clinical management in POTS patients. The thesis concludes that the different aspects of cardiorespiratory interactions can be modified acutely but the chronic effects of a long-term treatment or intervention on the magnitude of cardiorespiratory interactions are negligible and/or could be confused with the variability of markers. Considerations about the methodological dissimilarities and differences in effectiveness of the proposed indexes suggest that the simultaneous exploitation of all bivariate methodologies in cardiorespiratory studies is advantageous, as different aspects of cardiorespiratory interactions can be evaluated concurrently. This simultaneous evaluation can be carried out with a relatively negligible computational cost and in applicative contexts when only an ECG signal is available.File | Dimensione | Formato | |
---|---|---|---|
phd_unimi_R11992.pdf
Open Access dal 21/02/2022
Dimensione
1.56 MB
Formato
Adobe PDF
|
1.56 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/73650
URN:NBN:IT:UNIMI-73650