Home Button One Button Two Button Three Button Four

MEG/EEG

PET

Image Analysis

Optical Imaging

BrainStorm

BrainSuite

Aspire

Digimouse

MEG/EEG

PET

Image Analysis

Misc Topics

Group Members

Optical Tomography

 

In-vivo fluorescence and bioluminescence imaging plays a crucial role in interrogating biological systems, particularly, rodent models for human diseases. Majority of the systems to date generate only 2-D images of the integrated light distribution emitted from the surface of the animal, severely compromising the ability to perform quantify and accurately localize the signals due to the strong dependence of the tissue optical properties on depth. There is great interest in extending the in-vivo fluorescence and bioluminescence imaging modalities to the reconstruction of volumetric images that accurately localize signals and enable quantitative studies of fluorescent contrast agents and proteins. This is significant, as potentially higher temporal and spatial resolution than competing modalities like positron emission tomography can be achieved at a considerably low cost.

Incorporating spectral information about the fluorescent or bioluminescent molecules into the reconstruction process reduces the degree of illposedness of the problem of localizing the optical signals and therefore, increases the depth resolution dramatically than that achieved by traditional monochromatic data. Furthermore, multiple view angles i.e. tomographic acquisition of data contributes to enhance fluorescent or bioluminescent signal reconstructions. In in-vivo fluorescence imaging, spatially variant excitation illumination, potentially improves the conditioning of the inverse problem even further.

Figure: Principal component analysis of the spectral dependence on depth

 

In our work, we are developing the instrumentation and the algorithms that allows for tomographic data acquisition and reconstruction of three dimensional fluorescence or bioluminescence signal distributions in small animals. To achieve this goal, several issues are being addressed, including system design and calibration, designing of phantoms that simulate tissue, efficient and accurate modeling of light propagation in tissue, efficient image reconstruction strategies and a series of phantom and animal experiments that validate the approach.

 

Figure: Advantages of using hyperspectral information. Figure shows coronal sections through the bulk of the mouse taken at 2 mm steps and a simulated point source deep in tissue. The reconstruction using hyperspectral data is at the correct location