|
Quantitative analysis of 3D cerebral images in anatomical and functionnal MRI| old_uid | 9262 |
|---|
| title | Quantitative analysis of 3D cerebral images in anatomical and functionnal MRI |
|---|
| start_date | 2010/11/18 |
|---|
| schedule | 11h |
|---|
| online | no |
|---|
| location_info | salle du conseil |
|---|
| summary | Two different topics will be disussed during this talk, both of them being related to the quantitative analysis of 3D cerebral images: (1) Toward fine models for the quantitative analysis of the cerebral activity in fMRI. I will first present a model for the analysis of the cerebral activity in BOLD fMRI. Standard models make the hypothesis that the dynamic of oxygen consumption is homogeneous in the brain, although this consumption is obviously highly variable from a cortical area to another. Moreover, spatial smoothing of the images is performed in order to improve the signal to noise ratio. This however also leads to a loss of local activations and spatial accuracy in general. I will then present a model making finer hypotheses on the observed signal and discuss the gain of sensitivity it provides on clinical images. (2) Computational anatomy for the temporal analysis of the early brain development from MR images Computational anatomy (CA) is a mathematical formalism dedicated to the comparison of shapes with an unpreceded descriptive power. Although its mathematical foundations are well established, successful applications of this formalism on 3D images are very recent. After presenting the formalism, I will discuss its application to the anatomical analysis of the early brain development. 30 MR images were images acquired on pre-term babies between 30 and 40 weeks of gestational age. These images allowed us to build a model of the average cortical growth and its variability. I will then discuss our current results, and focus on the notion of scale at which the anatomical changes are observed. |
|---|
| responsibles | Fabre-Thorpe |
|---|
| |
|