My goal was to perform space and time mapping of blood flow in biological samples (Zebrafish embryos), being able to collect wide images (hundreds of microns), but still resolving them at a μm level; moreover hemodynamics is a fast process, that requires very high frequency sampling (tens to hundreds of Hz) in order to be resolved. The general biological motivation is that the progression of a wide number of diseases is affected directly by the blood circulation (an important example is cancer and its metastases), so that I want to devise an test methods to quantitatively map blood flow in different conditions, both for diagnosis and study of pathologies evolution . The approach I followed in my work is to employ Light Sheet Fluorescence Microscopy (LSFM) technique (also known as Selective-Plane Illumination Microscopy, SPIM) and a fast-acquisition, high efficiency camera, in order to achieve the minimum spatio-temporal resolution required (microns in space, 4ms in time). Acting on illumination profile, it is in fact possible to engineer a sheet of light to select just a thin (μm size) slice of the sample, so that fluorescence signals coming from just that plane can be measured in a 90° collection geometry. The CCD camera employed is an EMCCD, fabricated with a very advanced technology, capable of extremely high efficiency detection (quantum efficiency up to 95%), thus allowing very fast acquisition speeds. The last step was to fabricate samples and sample holder specifically designed to work with SPIM geometry, to achieve high transparency. Much work has been devoted to the study and manipulation of cross-correlation based algorithms, employed to retrieve flow parameters: being based on noise analysis, it is capable of excellent performances even in the most intricate biological situation I tried to investigate. The description of theoretical and practical aspects of my setup is the core of Chapter 2: I describe Single Plane Illumination Microscopy, focusing on a Physical Optics description of illumination and detection profiles. Then I will describe the practical implementation of two microscopes, one for in-vitro and one for in-vivo testing, showing an experimental evaluation of useful parameter derived in the previous section. The second part of the chapter is dedicated to the analysis of Cross-Correlation methods, providing both a solid presentation of all the techniques employed for image analysis, and also serving as an introduction for chapter 6, where an extension of these methods will be presented. Finally I will cover basic fluid dynamics to introduce a simple lumped circuit model, deriving the fundamental equation (Poiseuille flow in square channels) employed to describe in-vitro flows. Chapter 3 will describe material and methods, dealing with PDMS based microstructures, liposomes fabrication, and Zebrafish embryos description. In Chapter 4 I will summarize the results of the study of both in-vitro and in-vivo time-varying flows. Here I show a a first powerful application of temporal cross-correlation techniques coupled with large field of view images, which allows high resolution (both in time and space) mapping of flows. In Chapter 5 I will present related investigations, based on Image spatio-temporal correlations, in which I focused on in-vivo hemodynamics, in particular mapping blood flow in branched vessels in zebrafish embryos. Chapter 6 focuses on the most recent part of my work, that is to explore a way to break the "plane restriction" that seems to be intrinsically present when employing SPIM based microscopy. I will show that correlative methods can be extended allowing to retrieve 3D flow information, without any change in the hardware, as happens, for example, in optical tomography or micro-PIV. Finally Chapter 7 is dedicated to conclusions and future outlook.
Il mio principale obiettivo è stato effettuare una mappatura spazio-temporale del flusso sanguigno in campioni biologici (embrioni Zebrafish), riuscendo a registrare immagini a campo largo (centinaia di micron), pur risolvendole a livello micrometrico; inoltre, essendo l’emodinamica un processo molto rapido, richiede alta frequenza di campionamento (da decine a centinaia di Hertz). La motivazione biologica consiste nel fatto che la progressione di un gran numero di malattie è legata direttamente alla circolazione sanguigna (un esempio importante è rappresentato dal cancro con le sue metastasi), sicché ho voluto elaborare un metodo per mappare quantitativamente il flusso in differenti condizioni, sia a scopo diagnostico, sia per studiare l’evoluzione di patologie (ad esempio la risposta infiammatoria). L’approccio seguito consiste nell’impiego della tecnica LSFM (Light Sheet Fluorescence Microscopy) accoppiata ad una telecamera ad alta efficienza e rapida frequenza di acquisizione, per raggiungere la risoluzione spazio-temporale prefissata (micron nello spazio, 4ms nel tempo). Agendo sul profilo di illuminazione è possibile manipolare un fascio laser per creare un sottile (spessore di pochi μm) foglietto di luce che selezioni otticamente il campione: il segnale di fluorescenza può essere quindi raccolto con una geometria a 90°. La camera utilizzata è una EMCCD, capace di una grande efficienza di raccolta (efficienza quantica del 95%), e che permette rapide acquisizioni. L’ultimo passo è stata la costruzione di campioni e strutture create appositamente per lavorare con la geometria propria del microscopio LSFM e per raggiungere la massima trasparenza. Molto lavoro è stato dedicato alla manipolazione di algoritmi di cross-correlazione, impiegati per studiare parametri di flusso: questi metodi sono in grado di raggiungere prestazioni eccellenti anche nella più intricata situazione incontrata nell’analisi di campioni biologici. La descrizione degli aspetti teorici e pratici del mio setup è al centro del Capitolo 2: qui descrivo la teoria della LSFM, concentrandomi su una descrizione dei profili di eccitazione e raccolta basata sul modello dell’Ottica Fisica. In seguito illustro l’implementazione pratica di due microscopi, uno per misure in vitro, l’altro per misure in-vivo, mostrando una calibrazione sperimentale dei parametri descritti. La seconda parte del capitolo è dedicata all’analisi dei metodi di cross-correlazione, fornendo una solida presentazione di tutte le tecniche usate per l’analisi di immagini. Infine descrivo concetti chiave di dinamica dei fluidi arrivando ad introdurre un modello circuitale e derivando l’equazione fondamentale (flusso di Poiseuille in canaletti a sezione rettangolare) impiegato per descrivere i sistemi in-vitro. Il capitolo 3 descrive materiali e metodi, concentrandosi sulle microstrutture in PDMS, sulla fabbricazione di liposomi, e sulla descrizione degli embrioni di zebrafish. Nel capitolo 4 riassumo i risultati di uno studio (sia in-vitro che in-vivo) sui flussi variabili nel tempo. Qui mostro una prima applicazione delle tecniche di cross-correlazione temporale, applicate allo studio di immagini a campo largo, che permette una mappatura dei flussi con alta risoluzione spazio-temporale. Nel capitolo 5 presento studi basati su correlazione spazio-temporali, applicati all’emodinamica in-vivo, in particolare riuscendo a caratterizzare il complesso flusso in regioni con ramificazioni di vasi. Il capitolo 6 si concentra sulla parte più recente del mio lavoro, cioè la ricerca di un modo per rompere la “restrizione planare” che sembra essere intrinsecamente presente quando si impiega un sistema LSFM. Mostro che metodi correlativi possono essere estesi permettendo di ricavare informazioni su regimi di flusso tridimensionali, senza alcuna modifica nel setup. Infine il capitolo 7 è dedicato alle conclusioni ed alle prospettive future.
(2018). MICROFLUIDIC FLOW MAPPING WITH SPIM-ICS. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2018).
MICROFLUIDIC FLOW MAPPING WITH SPIM-ICS
CEFFA, NICOLÒ GIOVANNI
2018
Abstract
My goal was to perform space and time mapping of blood flow in biological samples (Zebrafish embryos), being able to collect wide images (hundreds of microns), but still resolving them at a μm level; moreover hemodynamics is a fast process, that requires very high frequency sampling (tens to hundreds of Hz) in order to be resolved. The general biological motivation is that the progression of a wide number of diseases is affected directly by the blood circulation (an important example is cancer and its metastases), so that I want to devise an test methods to quantitatively map blood flow in different conditions, both for diagnosis and study of pathologies evolution . The approach I followed in my work is to employ Light Sheet Fluorescence Microscopy (LSFM) technique (also known as Selective-Plane Illumination Microscopy, SPIM) and a fast-acquisition, high efficiency camera, in order to achieve the minimum spatio-temporal resolution required (microns in space, 4ms in time). Acting on illumination profile, it is in fact possible to engineer a sheet of light to select just a thin (μm size) slice of the sample, so that fluorescence signals coming from just that plane can be measured in a 90° collection geometry. The CCD camera employed is an EMCCD, fabricated with a very advanced technology, capable of extremely high efficiency detection (quantum efficiency up to 95%), thus allowing very fast acquisition speeds. The last step was to fabricate samples and sample holder specifically designed to work with SPIM geometry, to achieve high transparency. Much work has been devoted to the study and manipulation of cross-correlation based algorithms, employed to retrieve flow parameters: being based on noise analysis, it is capable of excellent performances even in the most intricate biological situation I tried to investigate. The description of theoretical and practical aspects of my setup is the core of Chapter 2: I describe Single Plane Illumination Microscopy, focusing on a Physical Optics description of illumination and detection profiles. Then I will describe the practical implementation of two microscopes, one for in-vitro and one for in-vivo testing, showing an experimental evaluation of useful parameter derived in the previous section. The second part of the chapter is dedicated to the analysis of Cross-Correlation methods, providing both a solid presentation of all the techniques employed for image analysis, and also serving as an introduction for chapter 6, where an extension of these methods will be presented. Finally I will cover basic fluid dynamics to introduce a simple lumped circuit model, deriving the fundamental equation (Poiseuille flow in square channels) employed to describe in-vitro flows. Chapter 3 will describe material and methods, dealing with PDMS based microstructures, liposomes fabrication, and Zebrafish embryos description. In Chapter 4 I will summarize the results of the study of both in-vitro and in-vivo time-varying flows. Here I show a a first powerful application of temporal cross-correlation techniques coupled with large field of view images, which allows high resolution (both in time and space) mapping of flows. In Chapter 5 I will present related investigations, based on Image spatio-temporal correlations, in which I focused on in-vivo hemodynamics, in particular mapping blood flow in branched vessels in zebrafish embryos. Chapter 6 focuses on the most recent part of my work, that is to explore a way to break the "plane restriction" that seems to be intrinsically present when employing SPIM based microscopy. I will show that correlative methods can be extended allowing to retrieve 3D flow information, without any change in the hardware, as happens, for example, in optical tomography or micro-PIV. Finally Chapter 7 is dedicated to conclusions and future outlook.File | Dimensione | Formato | |
---|---|---|---|
phd_unimib_702101.pdf
accesso aperto
Descrizione: tesi di dottorato
Tipologia di allegato:
Doctoral thesis
Dimensione
2.53 MB
Formato
Adobe PDF
|
2.53 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.