Introduction The ocular surface provides a unique opportunity to study hemodynamics since the vessels can be visualized directly, without treatment and non-invasively. The availability of instruments to measure various hemodynamic parameters on the ocular surface in an objective manner are lacking. The quantification of red blood cell velocity, blood oxygen saturation and conjunctival redness on the ocular surface using novel, validated techniques has the potential of providing useful information about vascular physiology. The specific aims of each chapter are as follows: Chapter 3: The objective was to design, develop and validate a system that would non-invasively quantify the red blood cell velocity in the conjunctival vessels. A tool was developed to automatically analyze video sequences of conjunctival vessels, digitally imaged with high enough magnification to resolve movement of the blood within the vessel. Chapter 4: The objective was to: a) design and develop a method in order to non-invasively quantify the changes in blood oxygen saturation (SO2) in the conjunctival vessels and demonstrate reliability of the measures and, b) demonstrate the application of the method by showing a response to an isocapnic hyperoxic provocation and compare those values to the results from a valid instrument. Chapter 5: The aim of this experiment was to examine variations in ocular redness levels, red blood cell velocities and oxygen saturation levels over time in clinically healthy participants and also to compare differences between two age groups. Chapter 6: The aim of this experiment was to examine the ocular redness levels, red blood cell velocities and oxygen saturation levels in clinically healthy participants when a topical ophthalmic decongestant was instilled onto the eye and to demonstrate the validity of the use of two novel techniques. Chapter 7: The aim of this experiment was to examine ocular redness, red blood cell velocity and oxygen saturation in participants who were habitual soft contact lens wearers (study) compared to those that did not (control) and also to compare differences in silicone (SH) and non-silicone hydrogel wearers. Methods Chapter 3: Simulations representing moving RBCs within a vessel and the random variation of each cell in terms of speed, shape and intensity were created in order to evaluate the performance of the algorithm. For each vessel, a signal that correlated to blood cell position was extracted from each frame, and the inter-frame displacement was estimated through a modified dynamic time warping (DTW) algorithm. This provided the red blood cell velocity over time in each point of the vessels. Thus, from these estimates, the mean red blood cell velocity for each vessel was easily evaluated. The true mean velocity from the simulation with the one estimated by the algorithm was compared and the system accuracy was determined. Chapter 4: a) Conjunctival vessels were imaged with two narrow-band interference filters with O2-sensitive and O2-insensitive peak transmissions using a Zeiss slit lamp at 32x magnification. Optical densities were calculated from vascular segments using the average reflected intensities inside and outside the vessels. Optical density ratios were used to calculate relative oxygen saturation values. Video images of the bulbar conjunctiva were recorded at three times of the day. Measurement repeatability was assessed over location at each time and across consecutive frames. b) Subjects initially breathed air for 10 minutes followed by pure oxygen (O2) for 20 minutes, and then air for a final 10 minute period using a sequential re-breathing circuit. Simultaneously, SO2 values measured with a pulse oximeter ear clip and finger clip were recorded. The validity of the dual wavelength method was demonstrated by comparing the values to those from the ear clip pulse oximeter. Chapter 5: Participants attended eight separate visits over the course of a day. Levels of bulbar conjunctival redness, red blood cell velocity and blood oxygen saturation were measured on a vessel of interest. Chapter 6: Participants attended three separate visits during an allotted 60 minute session. Bulbar conjunctival redness, red blood cell velocity and blood oxygen saturation were measured on a vessel of interest, pre-insertion, just after insertion and, 10 minutes after insertion of a topical ocular decongestant. Significant differences between the three measures were assessed and correlations between the three parameters were reported. Chapter 7: Participants were measured 8 times over the course of a day with their contact lenses in place. Bulbar conjunctival redness, red blood cell velocity and blood oxygen saturation were measured. Results Chapter 3: Results for the simulated videos demonstrated a very good concordance between the estimated and actual velocities supporting its validity. The mean relative error for the modified Dynamic Time Warping (DTW) method is 6%. Chapter 4: The intraclass correlations (ICCs) between the three locations at each time point were 0.93, 0.56 and 0.86 respectively. Measurements across 5 consecutive frames showed no significant difference for all subjects (ICC = 0.96). The ICCs between the two methods at each time point were 0.45, 0.10 and 0.11 respectively. a) There was no significant difference in SO2 between the three locations measured using the dual wavelength method for all subjects. There was also no significant difference between the three locations at any of the time points for the dual wavelength method. b) In response to isocapnic hyperoxic provocation using the dual wavelength method, blood oxygen saturation was increased from control values and subsequently recovered after withdrawal of hyperoxia. Blood oxygen saturation values recorded from the ear clip and finger clip of the pulse oximeter also showed an increase from control values and subsequently recovered after withdrawal of hyperoxia. SO2 comparison between the dual wavelength method and the ear-clip pulse oximeter method did not show a significant difference. The interaction between the two methods and time on SO2 was not significant. Chapter 5: From baseline, the group mean redness and oxygen saturation did not change significantly over time. There was a significant difference in the group mean red blood cell velocity values over time. There was no significant difference between age strata for all three measures. Chapter 6: After drop instillation redness values decreased significantly. There was no change in red blood cell velocity and oxygen saturation over time. There was a moderate significant correlation between SO2 and red blood cell velocity just after drop insertion. Chapter 7: When comparing the study and control groups, no significant difference in redness or SO2 over time was found. RBC velocity over time was found to be significantly different between groups. When comparing the two study groups (SH vs. hydrogel) no significant difference across either measure over time was found. Conclusions Chapter 3: Signal displacement estimation through the DTW algorithm can be used to estimate mean red blood cell velocity. Successful application of the algorithm in the estimation of RBC velocity in conjunctival vessels was demonstrated. Chapter 4: The application of the dual wavelength method was demonstrated and optical density ratios can be used in a reliable manner for relative oxygen saturation measurements. This valid method promises to enable the study of conjunctival O2 saturation under various experimental and physiological conditions. Chapter 5: The results of this study support the theory of metabolic regulation. The lack of any significant change across time for redness and oxygen saturation along with significant changes in red blood cell velocity substantiates this notion. Chapter 6: This study supports the literature regarding metabolic regulation of the microvasculature during the use of various stimuli. The results demonstrated that oxygen saturation levels remain stable even when a significant decrease in ocular redness is measured. The novel techniques used in this experiment demonstrated the expected action of the decongestant further contributing to their application and validity. Chapter 7: In summary, the participants in the study group were habitual contact lens wearers that had lower RBC velocities when compared to the control group supporting the notion that contact lenses initiate a hypoxic response. The lack of change in SO2 in either group supports the theory of metabolic regulation.

Introduction The ocular surface provides a unique opportunity to study hemodynamics since the vessels can be visualized directly, without treatment and non-invasively. The availability of instruments to measure various hemodynamic parameters on the ocular surface in an objective manner are lacking. The quantification of red blood cell velocity, blood oxygen saturation and conjunctival redness on the ocular surface using novel, validated techniques has the potential of providing useful information about vascular physiology. The specific aims of each chapter are as follows: Chapter 3: The objective was to design, develop and validate a system that would non-invasively quantify the red blood cell velocity in the conjunctival vessels. A tool was developed to automatically analyze video sequences of conjunctival vessels, digitally imaged with high enough magnification to resolve movement of the blood within the vessel. Chapter 4: The objective was to: a) design and develop a method in order to non-invasively quantify the changes in blood oxygen saturation (SO2) in the conjunctival vessels and demonstrate reliability of the measures and, b) demonstrate the application of the method by showing a response to an isocapnic hyperoxic provocation and compare those values to the results from a valid instrument. Chapter 5: The aim of this experiment was to examine variations in ocular redness levels, red blood cell velocities and oxygen saturation levels over time in clinically healthy participants and also to compare differences between two age groups. Chapter 6: The aim of this experiment was to examine the ocular redness levels, red blood cell velocities and oxygen saturation levels in clinically healthy participants when a topical ophthalmic decongestant was instilled onto the eye and to demonstrate the validity of the use of two novel techniques. Chapter 7: The aim of this experiment was to examine ocular redness, red blood cell velocity and oxygen saturation in participants who were habitual soft contact lens wearers (study) compared to those that did not (control) and also to compare differences in silicone (SH) and non-silicone hydrogel wearers. Methods Chapter 3: Simulations representing moving RBCs within a vessel and the random variation of each cell in terms of speed, shape and intensity were created in order to evaluate the performance of the algorithm. For each vessel, a signal that correlated to blood cell position was extracted from each frame, and the inter-frame displacement was estimated through a modified dynamic time warping (DTW) algorithm. This provided the red blood cell velocity over time in each point of the vessels. Thus, from these estimates, the mean red blood cell velocity for each vessel was easily evaluated. The true mean velocity from the simulation with the one estimated by the algorithm was compared and the system accuracy was determined. Chapter 4: a) Conjunctival vessels were imaged with two narrow-band interference filters with O2-sensitive and O2-insensitive peak transmissions using a Zeiss slit lamp at 32x magnification. Optical densities were calculated from vascular segments using the average reflected intensities inside and outside the vessels. Optical density ratios were used to calculate relative oxygen saturation values. Video images of the bulbar conjunctiva were recorded at three times of the day. Measurement repeatability was assessed over location at each time and across consecutive frames. b) Subjects initially breathed air for 10 minutes followed by pure oxygen (O2) for 20 minutes, and then air for a final 10 minute period using a sequential re-breathing circuit. Simultaneously, SO2 values measured with a pulse oximeter ear clip and finger clip were recorded. The validity of the dual wavelength method was demonstrated by comparing the values to those from the ear clip pulse oximeter. Chapter 5: Participants attended eight separate visits over the course of a day. Levels of bulbar conjunctival redness, red blood cell velocity and blood oxygen saturation were measured on a vessel of interest. Chapter 6: Participants attended three separate visits during an allotted 60 minute session. Bulbar conjunctival redness, red blood cell velocity and blood oxygen saturation were measured on a vessel of interest, pre-insertion, just after insertion and, 10 minutes after insertion of a topical ocular decongestant. Significant differences between the three measures were assessed and correlations between the three parameters were reported. Chapter 7: Participants were measured 8 times over the course of a day with their contact lenses in place. Bulbar conjunctival redness, red blood cell velocity and blood oxygen saturation were measured. Results Chapter 3: Results for the simulated videos demonstrated a very good concordance between the estimated and actual velocities supporting its validity. The mean relative error for the modified Dynamic Time Warping (DTW) method is 6%. Chapter 4: The intraclass correlations (ICCs) between the three locations at each time point were 0.93, 0.56 and 0.86 respectively. Measurements across 5 consecutive frames showed no significant difference for all subjects (ICC = 0.96). The ICCs between the two methods at each time point were 0.45, 0.10 and 0.11 respectively. a) There was no significant difference in SO2 between the three locations measured using the dual wavelength method for all subjects. There was also no significant difference between the three locations at any of the time points for the dual wavelength method. b) In response to isocapnic hyperoxic provocation using the dual wavelength method, blood oxygen saturation was increased from control values and subsequently recovered after withdrawal of hyperoxia. Blood oxygen saturation values recorded from the ear clip and finger clip of the pulse oximeter also showed an increase from control values and subsequently recovered after withdrawal of hyperoxia. SO2 comparison between the dual wavelength method and the ear-clip pulse oximeter method did not show a significant difference. The interaction between the two methods and time on SO2 was not significant. Chapter 5: From baseline, the group mean redness and oxygen saturation did not change significantly over time. There was a significant difference in the group mean red blood cell velocity values over time. There was no significant difference between age strata for all three measures. Chapter 6: After drop instillation redness values decreased significantly. There was no change in red blood cell velocity and oxygen saturation over time. There was a moderate significant correlation between SO2 and red blood cell velocity just after drop insertion. Chapter 7: When comparing the study and control groups, no significant difference in redness or SO2 over time was found. RBC velocity over time was found to be significantly different between groups. When comparing the two study groups (SH vs. hydrogel) no significant difference across either measure over time was found. Conclusions Chapter 3: Signal displacement estimation through the DTW algorithm can be used to estimate mean red blood cell velocity. Successful application of the algorithm in the estimation of RBC velocity in conjunctival vessels was demonstrated. Chapter 4: The application of the dual wavelength method was demonstrated and optical density ratios can be used in a reliable manner for relative oxygen saturation measurements. This valid method promises to enable the study of conjunctival O2 saturation under various experimental and physiological conditions. Chapter 5: The results of this study support the theory of metabolic regulation. The lack of any significant change across time for redness and oxygen saturation along with significant changes in red blood cell velocity substantiates this notion. Chapter 6: This study supports the literature regarding metabolic regulation of the microvasculature during the use of various stimuli. The results demonstrated that oxygen saturation levels remain stable even when a significant decrease in ocular redness is measured. The novel techniques used in this experiment demonstrated the expected action of the decongestant further contributing to their application and validity. Chapter 7: In summary, the participants in the study group were habitual contact lens wearers that had lower RBC velocities when compared to the control group supporting the notion that contact lenses initiate a hypoxic response. The lack of change in SO2 in either group supports the theory of metabolic regulation.

Au cours du développement du cancer, la migration des cellules cancéreuse en 3D joue un rôle essentiel dans le processus de dissémination des métastases. L’étude de la migration cellulaire dans des matrices 3D ainsi que les conséquences induites sur cette matrice sont actuellement étudiées par plusieurs équipes de recherche. Notamment, la réorganisation de la matrice extracellulaire et plus précisément les déplacements des fibres de la matrice induits par les forces que la cellule exerce sont des études en plein essor. Nous avons étudié comment les cellules cancéreuses migrent dans des gels 3D en utilisant du collagène et de la fibronectine pour mimer la matrice extracellulaire des tissus. Nous avons utilisé un microscope confocal afin de visualiser le cytosquelette d’actine des cellules en fluorescence et les fibres de collagène en réflexion. Dans ce travail,nous avons utilisé différentes concentrations de collagène et des lignées cellulaires d’invasivités différentes. A partir des films 3D obtenus en microscopie, nous avons déterminé la vitesse et la persistance des cellules cancéreuses en fonction de leur invasivité et de la concentration de collagène. La vitesse augmente avec l’invasivité cellulaire et diminue avec l’augmentation de la concentration en collagène. La persistance ne dépend que de la concentration en collagène et décroit avec celle-ci. Nous avons également calculé les champs de déplacement des fibres de collagène à l’aide d’un programme de corrélation de volume. Nous avons pu étudier ces champs de déplacement en fonction du type de migration de la cellule, de l’invasivité cellulaire et de la concentration en collagène des gels. Nous avons montré que les normes de vecteurs de déplacement augmentent avec l’invasivité cellulaire et diminuent avec l’augmentation de concentration en collagène. Enfin, ces champs de déplacement nous ont permis de déterminer les étapes des migrations mésenchymateuse et amiboïde en 3D. Nous avons découvert 5 étapes pour la migration mésenchymateuse correspondant au repos de la cellule, à la création d’une extension membranaire, à l’adhésion de la cellule aux fibres, au détachement de l’arrière du corps cellulaire afin de permettre à la cellule de migrer et à la dissolution de l’adhésion cellule/fibre. 4 étapes ont été déterminées pour la migration amiboïde et correspondent au repos de la cellule, à la création d’une extension membranaire, au déplacement de la cellule en poussant sur son environnement et à la rotation de la cellule. Ces étapes associées à des champs de déplacement sont en accord avec la littérature et nous avons pu mettre en évidence de nouvelles étapes comme la rotation de la cellule dans la migration amiboïde.Ces résultats permettent de mieux comprendre comment se déroule la migration des cellules cancéreuses dans une matrice extracellulaire. / 3D migration of cancer cells plays an essential role in the dissemination of cells during metastasisin cancer. The behavior of cancer cells migrating in a 3D extracellular matrix and its consequences on themicroenvironment are still currently under investigation. The study of the reorganization of the extracellular matrixfibers and more precisely how the fibers move due to the forces that the cell exerts just start to be investigating.We studied how cancer cells migrate in 3D gels using collagen and fibronectin to mimic the extracellularmatrix. We used confocal microscopy to image the actin cytoskeleton of cells in fluorescence and fibers in reflectionover time. In our studies, we used different collagen concentrations and cell lines with different invasivities. Fromthese 3D movies, we determined cancer cell velocities and persistence as a function of collagen gel concentration aswell as cell invasiveness. The cells velocities increase with invasiveness and decrease with collagen concentration.As for persistence, it decreases with collagen concentration but it do not change with cells invasiveness. We alsocalculated the displacement field of the collagen using a volume correlation program. Using this information, westudied the fibers displacement induced by the cell depending on its migration type, its invasivity and the collagenconcentration. We showed norms of fibers deplacement vectors increase with cell invasiveness and decrease withcollagen concentration. Finally, the displacement fields enabled us to determine the migration steps of mesenchymaland amiboid migrations. We discovered 5 steps in mesenchymal migration : cell rest, creation of extension, adhesionof the cell to the fibers, detachment of the cell rear and dissolution of cell/fibers adhesions. 4 steps have beencharacterized in amiboid migration : cell rest, creation of extension, displacement of the cell by pushing on fibersand rotation of the cell. These steps associated with displacement fields are in agreement with litterature and wehighlighted new steps as the rotation of the cell in amiboid migration.Taken together these results enable us to better understand how the migration of cancer cells takes place in a3D matrix.

Plusieurs décennies de recherche ont permis de mieux comprendre les effets de l’athérosclérose sur le système cardiovasculaire, d’améliorer la prévention et de développer des traitements efficaces. Les effets de l’athéroslérose sur le cerveau demeurent toutefois mal compris même si le lien entre le fonctionnement cognitif et la santé du système vasculaire est maintenant bien établi. La venue de nouvelles méthodes d’imagerie telle la microscopie laser à 2-photons (TPLM) permet d’étudier l’impact de certaines maladies sur la microvasculature cérébrale en mesurant le flux sanguin dans des vaisseaux uniques situés dans des régions cérébrales millimétriques sous la surface. Les résultats des études in vitro peuvent dorénavant être corrélés à ceux obtenus in vivo. En premier lieu, ce mémoire revoit la théorie ayant permis le développement de la TPLM qui permet de prendre des mesures hémodynamiques in vivo dans des vaisseaux de très petits calibres tels des capillaires cérébraux de souris. Par la suite, son utilisation est décrite chez des souris anesthésiées afin de comparer les mesures d’hémodynamie cérébrale tels la vitesse des globules rouges, le flux de globules rouges, le flux sanguin cérébral, l’hématocrite sanguin et le diamètre des vaisseaux. Finalement, nous avons comparé les données hémodynamiques entre des souris de 3 mois normales (WT ; n=6) et des souris atteintes d’athérosclérose précoce (ATX ; n=6). Les résultats obtenus sur un nombre total de 209 capillaires (103 pour les souris WT et 106 pour les souris ATX) démontrent que les souris ATX possèdent une vitesse des globules rouges (+40%) plus grande, un flux de globule rouge plus grand (+12%) et un flux capillaire plus élevé (+14%) sans démontrer pour aucun de ces paramètres, une différence statistiquement significative. L’hématocrite moyen (35±4% vs 33±2% ; p=0.71) et le diamètre moyen des vaisseaux (4.88±0.22μm vs 4.86±0.20μm ; p=0.23) étaient également comparables. La vitesse des globules rouges a démontré une faible corrélation avec le diamètre des vaisseaux (r=0.39) et avec le flux de globules rouges/seconde (r=0.59). En conclusion, les travaux menés dans le cadre de ce mémoire de maîtrise permettent d'envisager, grâce aux nouvelles méthodes d’imagerie cérébrale telle la TPLM, une meilleure compréhension des mécanismes hémodynamiques sous-jacents à la microcirculation cérébrale. L’effet d’une pression pulsée augmentée, tel que proposée dans l’athérosclérose reste cependant à démontrer avec cette méthode d’imagerie. / Many decades of research have given us a better understanding of the effects of atherosclerosis on different organs. For example, works on the cardiovascular effects of atherosclerosis have improved the prevention, screening and treatment resulting in a better prognosis. The effects of atherosclerosis on cerebral vessels are misunderstood even if the link between brain function and cerebral perfusion is well known. With the improvements in brain imaging, there are more possibilities to better define the physiopathology of brain perfusion and the impact of disease on cerebral microvasculature. First, this work reviews the theory behind the development of two-photon laser microscopy (TPLM) and the measure of hemodynamics in small vessels such as cerebral capillaries in mice. We also describe how we measured cerebral hemodynamics in anesthetized mice: red blood cell speed, red blood cell flux, blood hematocrit, and vessel diameter. We compared results obtained between normal mice (WT ; n=6) and pathological mice (ATX ; n=6), all aged 3 months old. ATX mice are well recognized to develop early atherosclerosis and served as a model for high pulse pressure. We measured 209 cerebral capillaries (103 on WT mice and 106 on ATX mice). ATX mice tend to show a trend toward a higher red blood cell speed (+40%), higher red blood cell flux (+12%) and higher capillary flux (+14%) in ATX mice. Mean hematocrit (35±4% vs 33±2% ; p=0.71) and mean vessel diameter (4.88±0.22μm vs 4.86±0.20μm ; p=0.23) were not statistically different between both groups. Red blood cell speed showed a weak correlation with vessel diameter (r=0.39) and red blood cell flux (r=0.59). In conclusion, the TPLM should permit a better understanding of the effect of vascular disease in cerebral hemodynamics. However, the effect of high pulse pressure on cerebral microvasculature needs to be better defined.