Perfusion Imaging Analysis in MEDx

Quantitative indices of regional cerebral hemodynamics are useful for identifying and monitoring many neurological and psychiatric conditions. These indices may be useful for determining the appropriate level of intervention following stroke. The MEDx medical image processing system now includes a new functional imaging module to create pixel-by-pixel maps of cerebral blood volume (CBV), cerebral blood flow (CBF), and mean transit time (MTT) from multiple whole brain scans.

Hemodynamic parameters are derived by first fitting the concentration curve to a gamma variate function. Determination of CBF requires an arterial input function for deconvolution of the output (implemented according to the method of Østergaard; MRM, 1996 a,b). The Perfusion Module includes a unique automated procedure for constructing the input function based on intelligent identification of arterial pixels from the raw data (Morris,et al,. 2000). The module also includes simpler techniques for computing CBV from 'area under curve' or change in steady state signal (Speck et al, MRM, 1999). For an introduction to perfusion imaging, see our perfusion slideshow.

Functional Data Simulator

For researchers who are interested in probing the analysis algorithms in MEDx or elsewhere, we now offer a powerful function data simulator. The simulator is a highly interactive and intuitive graphical tool for constricting realistic curves of signal intensity vs. time at a single pixel and then extrapolating to volumetric data. Users can create 4-dimensional data sets in MEDx with known parameters and then run these data through an analysis to gauge the effect on their CBV, CBF, and MTT maps of noise, delay, or curve shape. The 4-D data simulator can also be used to create synthetic fMRI data for similar explorations of the effects of noise, timing, etc., on the results of that technique.
 

Perfusion Module Highlights
	Functional data simulator 3D parametric maps of CBF, CBV, MTT Multiple algorithms for calculating CBV, CBF, and MTT Easy access to low-level algorithmic parameters for advanced users Automated and supervised algorithms for determination of an arterial input curve Arterial curve storage in ASCII format for quick and easy repeated analysis of data Transformation of original data images from 'susceptibility' to concentration units Interactive display of resultant parametric maps				Interactive simultaneous display of concentration data and fitted curve at each voxel Quality control (chi-squared) maps to evaluate the 'goodness of fit' at every voxel Integration into MEDx allows for convenient loading & manipulating of images Self-contained features for pre-processing (e.g. motion correcting), analyzing (CBV, CBF...) and post-processing (e.g. 'Talairaching') dynamic MR data all within the Perfusion module Multiple vendor formats and DICOM supported HTML formatted output reports viewable in Netscape

Top