The main function of the respiratory system is to guarantee appropriate gas exchange which involves the uptake of oxygen and release of carbon dioxide. Ventilatory control system attempts to correct deficiencies of oxygen (hypoxemia) and/or excess of carbon dioxide (hypercarbia) by increasing both tidal volume and respiratory rate, that are the main variables of breathing pattern. Breathing frequency, therefore, is per se an important informative parameter of respiratory function that may be a key predictor of adverse events. Monitoring breathing frequency, could early identify patients at risk of developing respiratory and cardiac dysfunction, with high specificity. The need for accurate, objective methods to assess and monitor breathing frequency is thus evident, inside and outside the clinic. Monitoring tools should be less invasive and obtrusive as possible, easy to use and low cost to foster the autonomous use by patients and promote the dissemination and transfer of the technology to the clinical and healthcare practice. Systems proposed in the literature for breathing pattern monitoring during daily life activities do not address these requirements and are confined to research application. An emerging approach that could answers to the abovementioned needs is based on the measurement of chest wall movements related to breath, using inertial sensors. Assessment of breathing pattern by measuring chest wall displacement, or derived tissue changes, it is acknowledged as preferred monitoring method because does not interfere with the airways, being minimally invasive. Inertial sensors open the door to new opportunities to develop this approach in a low-cost, wearable and easy to use manner. The present thesis wants to offer new insights into breathing monitoring by using inertial sensors, trying to overcome limitations of this fields, such as high sensitivity to motion artefacts and limited validation in clinical population. A new wearable, modular and wireless device for long-term monitoring of breathing frequency has been designed, developed, implemented, tested and optimized. It is based on sensor fusion of data collected from accelerometer, gyroscope and magnetometer (Magnetic and inertial sensor Unit, MIMU) to compute complete quaternion-based orientation, and its variation over time. The device is composed by three MIMU units. Two units are placed on the thorax and on the abdomen, to take into account the two-degree-of-freedom model of the chest wall. The third MIMU unit is used to have a local reference system, recording body (non-breathing) motion, and works as logging node. The device is integrated into an acquisition platform, in which data recorded by the units are sent to a smartphone and stored into a remote server, for telemonitoring service. A new position-independent algorithm, based on quaternion component fusion by using principal component analysis, has been proposed and compared against approaches based on quaternion component selection. It has been tested on healthy subjects in static and semi-static conditions, and adapted to dynamic conditions (walking), allowing automatic breath-by-breath extraction of breathing pattern temporal parameters. Finally, the device and method developed have been applied and tested on clinical target population, namely Muscular Dystrophy (MD). It is characterized by progressive muscle-weakening and wasting conditions, that affect also respiratory muscles, leading to respiratory failure, the main cause of death in these patients. A pilot study to assess usability, acceptance and wearability of the device in 15 patients with Duchenne Muscular Dystrophy and Limb-girdle Muscular Dystrophy has been recently approved by the Italian Minister of Health. The main aims of the study were to validate the device in static conditions, to assess its feasibility of autonomous use during daily activities and ultimately to collect information about the need of further design improvements. Data of this pilot study are reported in these thesis, representing the first available results in the literature about breathing rate assessment, obtained by using inertial-based devices in patients with respiratory muscular weakness. The proposed device provided good performance in terms of accuracy estimation in static conditions, and demonstrated to be well tolerated and accepted by patients, being able to monitor breathing rate variations over long period, also during daily activities. The research activity described in this thesis represents a step forward the implementation of at home continuous breathing frequency monitoring in patients at high risk of developing respiratory dysfunction and failure. The main function of the respiratory system is to guarantee appropriate gas exchange which involves the uptake of oxygen and release of carbon dioxide. Ventilatory control system attempts to correct deficiencies of oxygen (hypoxemia) and/or excess of carbon dioxide (hypercarbia) by increasing both tidal volume and respiratory rate, that are the main variables of breathing pattern. Breathing frequency, therefore, is per se an important informative parameter of respiratory function that may be a key predictor of adverse events. Monitoring breathing frequency, could early identify patients at risk of developing respiratory and cardiac dysfunction, with high specificity. The need for accurate, objective methods to assess and monitor breathing frequency is thus evident, inside and outside the clinic. Monitoring tools should be less invasive and obtrusive as possible, easy to use and low cost to foster the autonomous use by patients and promote the dissemination and transfer of the technology to the clinical and healthcare practice. Systems proposed in the literature for breathing pattern monitoring during daily life activities do not address these requirements and are confined to research application. An emerging approach that could answers to the abovementioned needs is based on the measurement of chest wall movements related to breath, using inertial sensors. Assessment of breathing pattern by measuring chest wall displacement, or derived tissue changes, it is acknowledged as preferred monitoring method because does not interfere with the airways, being minimally invasive. Inertial sensors open the door to new opportunities to develop this approach in a low-cost, wearable and easy to use manner. The present thesis wants to offer new insights into breathing monitoring by using inertial sensors, trying to overcome limitations of this fields, such as high sensitivity to motion artefacts and limited validation in clinical population. A new wearable, modular and wireless device for long-term monitoring of breathing frequency has been designed, developed, implemented, tested and optimized. It is based on sensor fusion of data collected from accelerometer, gyroscope and magnetometer (Magnetic and inertial sensor Unit, MIMU) to compute complete quaternion-based orientation, and its variation over time. The device is composed by three MIMU units. Two units are placed on the thorax and on the abdomen, to take into account the two-degree-of-freedom model of the chest wall. The third MIMU unit is used to have a local reference system, recording body (non-breathing) motion, and works as logging node. The device is integrated into an acquisition platform, in which data recorded by the units are sent to a smartphone and stored into a remote server, for telemonitoring service. A new position-independent algorithm, based on quaternion component fusion by using principal component analysis, has been proposed and compared against approaches based on quaternion component selection. It has been tested on healthy subjects in static and semi-static conditions, and adapted to dynamic conditions (walking), allowing automatic breath-by-breath extraction of breathing pattern temporal parameters. Finally, the device and method developed have been applied and tested on clinical target population, namely Muscular Dystrophy (MD). It is characterized by progressive muscle-weakening and wasting conditions, that affect also respiratory muscles, leading to respiratory failure, the main cause of death in these patients. A pilot study to assess usability, acceptance and wearability of the device in 15 patients with Duchenne Muscular Dystrophy and Limb-girdle Muscular Dystrophy has been recently approved by the Italian Minister of Health. The main aims of the study were to validate the device in static conditions, to assess its feasibility of autonomous use during daily activities and ultimately to collect information about the need of further design improvements. Data of this pilot study are reported in these thesis, representing the first available results in the literature about breathing rate assessment, obtained by using inertial-based devices in patients with respiratory muscular weakness. The proposed device provided good performance in terms of accuracy estimation in static conditions, and demonstrated to be well tolerated and accepted by patients, being able to monitor breathing rate variations over long period, also during daily activities. The research activity described in this thesis represents a step forward the implementation of at home continuous breathing frequency monitoring in patients at high risk of developing respiratory dysfunction and failure.
La funzione del sistema respiratorio è quella di garantire un appropriato apporto di ossigeno ai tessuti e la rimozione dell’anidride carbonica dagli stessi. Il sistema di controllo della ventilazione opera quindi per correggere deficienze di ossigeno (ipossiemia) e/o eccessi di anidride carbonica (ipercapnia) mediante la regolazione del volume corrente e della frequenza respiratoria, due parametri fondamentali che caratterizzano il pattern respiratorio. La frequenza respiratoria è un parametro che racchiude importantissime informazioni sulla funzionalità respiratoria, ed è un indice chiave in grado di predire eventi clinici avversi. Con un accurato monitoraggio della frequenza respiratoria è possibile identificare con elevata specificità pazienti a rischio di sviluppare problematiche cardiache e respiratorie. Si fa evidente quindi la necessità di disporre di metodi oggettivi e accurati per la misura e il monitoraggio continuo di questo fondamentale parametro vitale. Per garantire l’utilizzo di tali strumenti da parte del paziente anche al di fuori della clinica e promuovere il trasferimento tecnologico di tali tecnologie alla pratica clinica, gli strumenti di monitoraggio dovrebbero essere, oltre che accurati, il meno invasivi possibile, facili da usare e a basso costo. I sistemi proposti in letteratura per il monitoraggio del pattern respiratorio, anche durante le attività giornaliere, non rispondono a questi requisiti e sono nella maggior parte dei casi confinati ad applicazioni di ricerca. Un approccio emergente che potrebbe invece trovare applicazione in questo campo si basa sulla misura, mediante sensori inerziali, dei movimenti della parete toracoaddominale dovuti all’attività respiratoria. La derivazione del segnale respiratorio da movimenti della superficie toracoaddominale o dai cambiamenti tissutali ad essi legati, è riconosciuto come metodo preferenziale per il monitoraggio respiratorio sul lungo periodo, perché non interferisce con le vie aeree ed è quindi minimamente invasivo ed intrusivo, permettendo il normale svolgimento delle attività quotidiane. I sensori inerziali aprono la strada allo sviluppo di sistemi economici, indossabili e semplici da usare. Il lavoro di ricerca di questa tesi si pone come obiettivo l’approfondimento di questo approccio, cercando di superare i limiti emersi in letteratura legati principalmente agli artefatti da movimento e alla limitata validazione in clinica. Viene presentato un nuovo dispositivo indossabile, modulare e wireless per il monitoraggio sul lungo periodo della frequenza respiratoria, descrivendo l’intero processo di sviluppo, dalla progettazione alla validazione. Il dispositivo si compone di tre unità sensorizzate MIMU. Ogni MIMU contiene accelerometro, un giroscopio e un magnetometro dalla cui fusione è possibile ottenere l’orientamento dell’unità nello spazio e le sue variazioni nel tempo. Due delle tre unità sensorizzate sono poste rispettivamente su torace e addome per la registrazione dei movimenti toracoaddominali respiratori; l’utilizzo di due unità permettere di tenere in considerazione il modello a due gradi di libertà della parete toracoaddominale, che vede torace e addome come compartimenti indipendenti. La terza unità è l’unità centrale del sistema: salva i dati e si interfaccia con uno smartphone che funziona da gateway verso un server web. Il device costituito dalle tre unità sensorizzate è quindi inserito in una piattaforma di acquisizione utilizzabile anche in servizi di tele-monitoraggio. Inoltre, questa unità contiene una MIMU e viene posizionata su una parte del corpo solidale al tronco ma non coinvolta in movimenti respiratori; in questo modo questa unità funge da sistema di riferimento locale registrando i movimenti del corpo. Oltre alla progettazione e realizzazione della parte hardware e firmware il mio lavoro di ricerca si è focalizzato anche sullo sviluppo di algoritmi di analisi dei dati raccolti. In particolare, in questa tesi propongo un algoritmo di analisi in grado di estrarre i parametri respiratori da segnali di movimento toracoaddominale, indipendentemente dalla postura assunta e dal posizionamento. Sono stati considerati diversi approcci di analisi, confrontando i risultati ottenuti in un gruppo di soggetti sani in condizioni statiche. L’approccio migliore è stato successivamente adattato ed esteso permettendo l’analisi del pattern respiratorio respiro per respiro in condizioni semi-statiche e dinamiche. Infine il dispositivo e il metodo sono stati applicati e testati in pazienti con distrofia muscolare, una patologia caratterizzata da un progressivo indebolimento muscolare. In determinate forme di distrofia muscolare anche i muscoli respiratori vengono indeboliti portando nel giro di pochi anni dall’esordio, a insufficienza respiratoria, la principale causa di morte in questi pazienti. Il Ministero della Salute Italiano ha approvato uno studio pilota per verificare l’usabilità, la vestibilità e l’accuratezza del dispositivo in 15 pazienti con distrofia muscolare (distrofia muscolare di Duchenne e distrofia muscolare dei cingoli). L’obiettivo principale dello studio è quello di validare il dispositivo in condizioni statiche e di verificare la fattibilità dell’utilizzo prolungato e autonomo da parte dei pazienti e dei caregivers durante le attività quotidiane. Infine, un aspetto fondamentale della sperimentazione che è attualmente in corso riguarda la possibilità di raccogliere informazioni circa la necessità di modifiche di progettazione del dispositivo. I primi dati preliminari dello studio sono riportati nella presente tesi e rappresentato i primi dati disponibili in letteratura riguardanti il monitoraggio respiratorio prolungato mediante sensori inerziali in pazienti con distrofia muscolare o con debolezza muscolare. Il dispositivo proposto ha mostrato ottimi risultati in termini di accuratezza nella stima della frequenza respiratoria in condizioni statiche e in termini di vestibilità, usabilità e capacità di monitorare le variazioni del tracciato respiratorio durante le attività quotidiane. L’attività di ricerca riportata nella presente tesi rappresenta quindi un importante passo avanti nell’implementazione di un sistema indossabile per il monitoraggio respiratorio continuo al domicilio in pazienti ad alto rischio di sviluppare problematiche respiratorie e insufficienza respiratoria.
RespirHó: an inertial sensor-based system for daily life breathing monitoring in neuromuscular patients
CESAREO, AMBRA
Abstract
The main function of the respiratory system is to guarantee appropriate gas exchange which involves the uptake of oxygen and release of carbon dioxide. Ventilatory control system attempts to correct deficiencies of oxygen (hypoxemia) and/or excess of carbon dioxide (hypercarbia) by increasing both tidal volume and respiratory rate, that are the main variables of breathing pattern. Breathing frequency, therefore, is per se an important informative parameter of respiratory function that may be a key predictor of adverse events. Monitoring breathing frequency, could early identify patients at risk of developing respiratory and cardiac dysfunction, with high specificity. The need for accurate, objective methods to assess and monitor breathing frequency is thus evident, inside and outside the clinic. Monitoring tools should be less invasive and obtrusive as possible, easy to use and low cost to foster the autonomous use by patients and promote the dissemination and transfer of the technology to the clinical and healthcare practice. Systems proposed in the literature for breathing pattern monitoring during daily life activities do not address these requirements and are confined to research application. An emerging approach that could answers to the abovementioned needs is based on the measurement of chest wall movements related to breath, using inertial sensors. Assessment of breathing pattern by measuring chest wall displacement, or derived tissue changes, it is acknowledged as preferred monitoring method because does not interfere with the airways, being minimally invasive. Inertial sensors open the door to new opportunities to develop this approach in a low-cost, wearable and easy to use manner. The present thesis wants to offer new insights into breathing monitoring by using inertial sensors, trying to overcome limitations of this fields, such as high sensitivity to motion artefacts and limited validation in clinical population. A new wearable, modular and wireless device for long-term monitoring of breathing frequency has been designed, developed, implemented, tested and optimized. It is based on sensor fusion of data collected from accelerometer, gyroscope and magnetometer (Magnetic and inertial sensor Unit, MIMU) to compute complete quaternion-based orientation, and its variation over time. The device is composed by three MIMU units. Two units are placed on the thorax and on the abdomen, to take into account the two-degree-of-freedom model of the chest wall. The third MIMU unit is used to have a local reference system, recording body (non-breathing) motion, and works as logging node. The device is integrated into an acquisition platform, in which data recorded by the units are sent to a smartphone and stored into a remote server, for telemonitoring service. A new position-independent algorithm, based on quaternion component fusion by using principal component analysis, has been proposed and compared against approaches based on quaternion component selection. It has been tested on healthy subjects in static and semi-static conditions, and adapted to dynamic conditions (walking), allowing automatic breath-by-breath extraction of breathing pattern temporal parameters. Finally, the device and method developed have been applied and tested on clinical target population, namely Muscular Dystrophy (MD). It is characterized by progressive muscle-weakening and wasting conditions, that affect also respiratory muscles, leading to respiratory failure, the main cause of death in these patients. A pilot study to assess usability, acceptance and wearability of the device in 15 patients with Duchenne Muscular Dystrophy and Limb-girdle Muscular Dystrophy has been recently approved by the Italian Minister of Health. The main aims of the study were to validate the device in static conditions, to assess its feasibility of autonomous use during daily activities and ultimately to collect information about the need of further design improvements. Data of this pilot study are reported in these thesis, representing the first available results in the literature about breathing rate assessment, obtained by using inertial-based devices in patients with respiratory muscular weakness. The proposed device provided good performance in terms of accuracy estimation in static conditions, and demonstrated to be well tolerated and accepted by patients, being able to monitor breathing rate variations over long period, also during daily activities. The research activity described in this thesis represents a step forward the implementation of at home continuous breathing frequency monitoring in patients at high risk of developing respiratory dysfunction and failure. The main function of the respiratory system is to guarantee appropriate gas exchange which involves the uptake of oxygen and release of carbon dioxide. Ventilatory control system attempts to correct deficiencies of oxygen (hypoxemia) and/or excess of carbon dioxide (hypercarbia) by increasing both tidal volume and respiratory rate, that are the main variables of breathing pattern. Breathing frequency, therefore, is per se an important informative parameter of respiratory function that may be a key predictor of adverse events. Monitoring breathing frequency, could early identify patients at risk of developing respiratory and cardiac dysfunction, with high specificity. The need for accurate, objective methods to assess and monitor breathing frequency is thus evident, inside and outside the clinic. Monitoring tools should be less invasive and obtrusive as possible, easy to use and low cost to foster the autonomous use by patients and promote the dissemination and transfer of the technology to the clinical and healthcare practice. Systems proposed in the literature for breathing pattern monitoring during daily life activities do not address these requirements and are confined to research application. An emerging approach that could answers to the abovementioned needs is based on the measurement of chest wall movements related to breath, using inertial sensors. Assessment of breathing pattern by measuring chest wall displacement, or derived tissue changes, it is acknowledged as preferred monitoring method because does not interfere with the airways, being minimally invasive. Inertial sensors open the door to new opportunities to develop this approach in a low-cost, wearable and easy to use manner. The present thesis wants to offer new insights into breathing monitoring by using inertial sensors, trying to overcome limitations of this fields, such as high sensitivity to motion artefacts and limited validation in clinical population. A new wearable, modular and wireless device for long-term monitoring of breathing frequency has been designed, developed, implemented, tested and optimized. It is based on sensor fusion of data collected from accelerometer, gyroscope and magnetometer (Magnetic and inertial sensor Unit, MIMU) to compute complete quaternion-based orientation, and its variation over time. The device is composed by three MIMU units. Two units are placed on the thorax and on the abdomen, to take into account the two-degree-of-freedom model of the chest wall. The third MIMU unit is used to have a local reference system, recording body (non-breathing) motion, and works as logging node. The device is integrated into an acquisition platform, in which data recorded by the units are sent to a smartphone and stored into a remote server, for telemonitoring service. A new position-independent algorithm, based on quaternion component fusion by using principal component analysis, has been proposed and compared against approaches based on quaternion component selection. It has been tested on healthy subjects in static and semi-static conditions, and adapted to dynamic conditions (walking), allowing automatic breath-by-breath extraction of breathing pattern temporal parameters. Finally, the device and method developed have been applied and tested on clinical target population, namely Muscular Dystrophy (MD). It is characterized by progressive muscle-weakening and wasting conditions, that affect also respiratory muscles, leading to respiratory failure, the main cause of death in these patients. A pilot study to assess usability, acceptance and wearability of the device in 15 patients with Duchenne Muscular Dystrophy and Limb-girdle Muscular Dystrophy has been recently approved by the Italian Minister of Health. The main aims of the study were to validate the device in static conditions, to assess its feasibility of autonomous use during daily activities and ultimately to collect information about the need of further design improvements. Data of this pilot study are reported in these thesis, representing the first available results in the literature about breathing rate assessment, obtained by using inertial-based devices in patients with respiratory muscular weakness. The proposed device provided good performance in terms of accuracy estimation in static conditions, and demonstrated to be well tolerated and accepted by patients, being able to monitor breathing rate variations over long period, also during daily activities. The research activity described in this thesis represents a step forward the implementation of at home continuous breathing frequency monitoring in patients at high risk of developing respiratory dysfunction and failure.File | Dimensione | Formato | |
---|---|---|---|
Submitted_PhD_Thesis_CesareoAmbra_13112018.pdf
solo utenti autorizzati dal 14/11/2021
Descrizione: Testo della tesi
Dimensione
10.37 MB
Formato
Adobe PDF
|
10.37 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/143668