Image Reconstruction
All imaging require reconstruction. The simplest, of course, are planar techniques, in which signal intensity of x,y coordinates is used directly to form the 2D image. Image reconstruction of MR images typically involves 2 or 3D Fourier transforms or less typically modified back projection techniques (which are usually employed in CT image reconstruction). Fluorescence-mediated molecular tomographic optical imaging uses iterative reconstruction, based on the diffusion equation.

Tissue concentrations and/or activation of imaging probe potentially provide information regarding the molecular expression within organs. Nuclear and MR imaging for example, are inherently quantitative. With different optical imaging technologies quantitation is either inherent (e.g. diffuse optical tomography) or more difficult (e.g reflectance optical imaging, bioluminescence imaging) because of light diffusion. However, “correlative quantitation”, ratio imaging and temporal imaging may still result in important biological information.

The quantitative results achieved by imaging are most useful when reproducible and validated by another means. The imaging systems/imaging paradigms designed at CMIR, are extensively calibrated by ex vivo techniques, including radiolabeled biodistribution assays, elemental analysis, and second fluorescence label quantitation.

Data Display, Fusion
Molecular information is most useful when combined with anatomic information which reveals where in the body the signal is originating from. Markers for aligning images acquired using different imaging systems is a technique currently employed at CMIR. Human imaging system manufacturers have also overcome this problem by making machines which can, for example, acquire SPECT and CT images nearly simultaneously.