Spatial Frequency Domain Imaging (SFDI) is a model-based, wide-field, non-contact method for measuring the absorption, scattering, and fluorescence properties of biological tissue. Optical properties are determined in each pixel simultaneously, by measuring the attenuation (or fluorescence) of sinusoidal patterns of light projected onto the sample at varying spatial frequencies and phases. Images are demodulated by processing 3 phase-shifted views of the sample. The mean interrogation depth at a given wavelength is controlled by the spatial frequency of projection, and frequency-dependent differences in path length are used to calculate tissue optical properties using computational models. Because 3 specific phases are required for each projected frequency, care must be taken to perfectly sequence all projections and camera triggers. While each of these processes is fairly rapid, together they can slow the acquisition to a fraction of the camera frame rate. In order to overcome this limitation and facilitate real-time SFDI, we will develop new methods using """"""""frequency synthesis"""""""" - multiple frequencies synthesized into customized projection patterns. These patterns will be optimized for speed and frequency-dependent information content in order to facilitate rapid and accurate optical property measurements, probe buried objects, and perform tomography. When properly selected, frequency synthesized projections can potentially decrease the minimum acquisition time to the frame rate of the camera, allowing real-time SFDI and SFD tomography. The ability to project custom patterns not only allows us to generate multi-frequency components, it also adds the ability to change their orientation. This allows us to explore a new mode of contrast based on probing tissue structures that are aligned with the direction of the projected pattern. This is due to the fact that many tissue types, including bone, muscle, skin, and white matter in the brain, have orientated internal structures such that the degree of optical scattering depends on the direction of light propagation. The scattering direction of these oriented tissues is determined by their microscopic structure and obeys a diffusion equation. We will derive accurate solutions to the anisotropic diffusion equation in the spatial frequency domain. In an ordered medium, the attenuation of sinusoidal patterns depends on the relative orientation of the spatial frequency pattern and scatterer direction. Thus, by projecting multiple spatial frequencies in different directions and measuring the attenuation, we will be able to image the spatially varying scattering orientation over a large field of view. We expect that the combination of spatial frequency synthesis and orientation control will lead to new methods for quantitative, real-time imaging and tomography in thick tissues, as well as the characterization of exciting new contrast mechanisms based on an optical diffusion tensor.
These studies will allow us to acquire sufficient data to develop and validate a new optical technology for real-time, quantitative imaging and depth sectioning of tissues based on relatively simple, cost-effective imaging cameras and patterned light projection techniques. The technology has potential to replace conventional camera-based imaging methods by allowing quantitative viewing of functional tissue attributes beneath the surface, where disease typically begins. Our approach is expected to lead to new, bedside medical imaging methods for detecting disease, monitoring therapy response, and guiding surgical procedures.
Showing the most recent 10 out of 11 publications
