Étude de la réponse hémodynamique dans un modèle réussi de vieillissement chez le rat par imagerie optique intrinsèque

The aim of this thesis is to measure changes in neurovascular parameters and in hemodynamic response in aging rats. The hemodynamic response is the process by which, following an increase of neuronal activity, the blood flow increase locally to follow the changes in metabolic activity from neural activity. Measuring the hemodynamic response is the key process of functional imaging and allows to monitor indirectly changes of neuronal activity through the neurovascular coupling. However, many physiological properties in the brain (flow, blood volume, vessel compliance, vascular density, etc.) may be modified during aging and affect neurovascular coupling. The thesis aims to better understand the changes in neurovascular coupling during aging and their effect measurements obtained in functional imaging.

In this work, changes in hemodynamic response were measured using an intrinsic optical imaging system (IOI) developed in the laboratory. This recent imaging technique is based on the absorption properties of the visible light in the cortex. The technique has a very good spatial resolution and low depth of penetration which makes it well suited to study brain activity in rats. In IOI, visible light illuminates the cortex and travels in the surface layer of the brain before being reflected at the surface of the cortex and measured using a CCD camera. In the journey through the cortex, a part of the light is absorbed by the two main chromophores present (hemoglobin and deoxyhemoglobin). Thus, changes in chromophore concentration can be determined through changes in light intensity. This allows to study the concentrations of oxygenated blood and deoxygenated. In addition to measures of IOI, the system simultaneously measures changes in blood flow through a laser speckle measurement technique.

The first part of the results then shows the changes in the hemodynamic response during aging. The main observations is that during aging we observe a decrease in the amplitude of the hemodynamic response and an increase in the activation time of the hemodynamic response. Age also produce changes of the hemodynamic response. Thus, the amplitude of the hemodynamic response decreases more slowly as a function of aging on the side ipsilateral to stimulation from the contralateral side.

The second part of the work studies neurovascular coupling using three different biophysical models. The three biophysical models can reproduce the different types of hemodynamic response found in our rats population. A comparison of models by log evidence did not foound significative difference between the performance of the three models. However, the model-Boas Huppert that we developed has an advantage of finding neurovascular resting state parameters. The work confirmed that the three biophysical models find neurovascular parameters with enough precision to do group comparisons. All three models have shown a decrease in compliance of vessels with age.

The third part of the work examines a recent model of deconvolution of the hemodynamic response based on Kalman filters. This deconvolution is used to estimate the neuronal activity for a complete recording sequence instead of working with averaged responses. The work shows that the deconvolution of the hemodynamic response gives a good estimated of neural activity when that activity is produced by external stimulation. When one focuses only on the spontaneous neuronal activity (at rest), we find that neuronal activity correlated weakly with neural activity measured by electrophysiology. This indicates that caution should be exercised in the interpretation that is made of spontaneous neuronal activity obtained by deconvolution.