Chronic respiratory diseases are among the most common causes of severe illness and death worldwide. An early and accurate diagnosis becomes crucial for a prompt and specific therapeutic intervention. However, despite being effective, the current diagnostic methods present limitations, as they do not provide affordable, rapid, and harmless solutions for accurate early detection of the diseases. In this context, the analysis of Volatile Organic Compounds (VOCs) in exhaled breath (breath-omics), represents a promising alternative solution, as it allows an affordable, effective, and non-invasive analysis of human exhaled breath. In the last years electronic noses (E-nose) revealed to be an encouraging technology to detect VOCs in exhaled breath and to correlate them with pathological conditions via machine learning algorithms. However, the lack of standardized approach for sampling and analyze exhaled breath hampers the pathway toward clinical applications. Moreover, several confounding factors that can interfere during collection and storage prior to analysis challenge obtaining reliable measurements. According to these requirements, this thesis project aims at examining the actual E-nose technology for exhaled breath analysis for early diagnosis of respiratory diseases by developing and testing a system suitable for conducting in-vivo preclinical trials, and by optimizing an E-nose setup for real-time breath analysis to overcome the above mentioned limitations. The exploratory analysis of an in-vivo dataset from experimental model showed potential for distinguishing between diseased and healthy subjects. However, it emphasized the importance of compensating for FiO2 and obtaining more samples for robust classification. The in-vitro tests demonstrated the feasibility of the E-nose setup for VOCs on-line analysis under simulated-breath conditions, highlighting the need for implementing humidity management.
Le malattie respiratorie croniche sono tra le cause più comuni di patologia grave e di morte in tutto il mondo. Una diagnosi precoce e accurata diventa cruciale per un intervento terapeutico immediato e specifico. Tuttavia, nonostante la loro validità, le attuali metodologie diagnostiche presentano limitazioni, poiché non forniscono soluzioni economiche, rapide e innocue per una diagnosi tempestiva e precisa delle malattie. In questo contesto, l'analisi dei composti organici volatili (VOC) presenti nell'esalato (breathomics) rappresenta una soluzione alternativa promettente, poiché consente di analizzare l'espirato in maniera non invasiva, efficace ed economica. Negli ultimi anni, i nasi elettronici si sono rivelati una tecnologia promettente per rilevare i VOC nell'esalato e per correlarli con le condizioni patologiche mediante algoritmi di apprendimento automatico (machine learning). Tuttavia, l'assenza di un approccio standardizzato per il campionamento e l'analisi del respiro ostacola il percorso verso l'implementazione clinica. Inoltre, i numerosi fattori confondenti che possono interferire durante la raccolta e il condizionamento precedenti all'analisi rappresentano una sfida per ottenere misurazioni affidabili. In base a questi requisiti, questo progetto di tesi mira ad esaminare la tecnologia del naso elettronico per l'analisi del respiro nella diagnosi precoce di malattie respiratorie, sviluppando un sistema adatto per condurre studi preclinici in-vivo e ottimizzando un prototipo di naso elettronico in grado di effettuare analisi in continua ed in tempo reale, per superare le limitazioni sopra menzionate. L'analisi esplorativa del dataset in-vivo ha mostrato un potenziale per distinguere tra soggetti malati e sani. Al contempo, ha sottolineato l'importanza della compensazione della frazione inspirata di ossigeno (FiO2) e dell'ottenimento di più campioni per una classificazione robusta. I test in-vitro hanno dimostrato la fattibilità della configurazione del naso elettronico per l'analisi on-line dei VOC in condizioni simil-esalato, evidenziando la necessità di implementare opportuni sistemi di gestione dell’umidità.
An electronic nose device for direct sampling of exhaled breath in patients with respiratory conditions
Pierantozzi, Aurora
2021/2022
Abstract
Chronic respiratory diseases are among the most common causes of severe illness and death worldwide. An early and accurate diagnosis becomes crucial for a prompt and specific therapeutic intervention. However, despite being effective, the current diagnostic methods present limitations, as they do not provide affordable, rapid, and harmless solutions for accurate early detection of the diseases. In this context, the analysis of Volatile Organic Compounds (VOCs) in exhaled breath (breath-omics), represents a promising alternative solution, as it allows an affordable, effective, and non-invasive analysis of human exhaled breath. In the last years electronic noses (E-nose) revealed to be an encouraging technology to detect VOCs in exhaled breath and to correlate them with pathological conditions via machine learning algorithms. However, the lack of standardized approach for sampling and analyze exhaled breath hampers the pathway toward clinical applications. Moreover, several confounding factors that can interfere during collection and storage prior to analysis challenge obtaining reliable measurements. According to these requirements, this thesis project aims at examining the actual E-nose technology for exhaled breath analysis for early diagnosis of respiratory diseases by developing and testing a system suitable for conducting in-vivo preclinical trials, and by optimizing an E-nose setup for real-time breath analysis to overcome the above mentioned limitations. The exploratory analysis of an in-vivo dataset from experimental model showed potential for distinguishing between diseased and healthy subjects. However, it emphasized the importance of compensating for FiO2 and obtaining more samples for robust classification. The in-vitro tests demonstrated the feasibility of the E-nose setup for VOCs on-line analysis under simulated-breath conditions, highlighting the need for implementing humidity management.File | Dimensione | Formato | |
---|---|---|---|
2023_Pierantozzi.pdf
non accessibile
Descrizione: manoscritto
Dimensione
8.57 MB
Formato
Adobe PDF
|
8.57 MB | Adobe PDF | Visualizza/Apri |
Executive_Summary_PierantozziAurora.pdf
non accessibile
Descrizione: executive summary
Dimensione
1.11 MB
Formato
Adobe PDF
|
1.11 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/10589/202921