Near-Infrared (NIR) spectroscopic tomography has been emerging as a method to image blood-based contrast in tissue. A potential strength of this imaging approach which has not been fully exploited to date is that it can be very fast, essentially real-time, similar to ultrasound. A major limitation in the development of fast NIR tomography systems has been the need to sequentially scan the light source, such that even with parallel detection, the image acquisition time is limited to frame rates of a few Hertz. This work introduces a new concept for imaging in parallel with all source locations simultaneously activated. The technique uses an array of light sources that are separated by fractions of a nanometer spectrally, yet clustered within a narrow (4 to 5 nm) bandwidth. Since the spectral features of tissue do not vary significantly over such a narrow band, these different laser sources will sample the same attenuation and scattering processes in tissue, yet in the detection channel the response associated with each source wavelength can readily be separated by a spectrograph, and thereby detected in parallel. This approach avoids the problem of source saturation of the detectors, while allowing parallel detection of all sources and all detectors in a single measurement event. This type of spectral encoding of the position is at the heart of several imaging systems such as MRI, where it allows parallel readout of the location from different radio frequencies. The system will be implemented in a small animal imaging geometry, using CCD and photomultiplier tube detection. Feasibility will be demonstrated in hardware and software through both tissue phantom and animal experiments. The hemodynamic response in terms of pulsatile flow in the brain will be imaged in the rat brain, as well as indocyanine green kinetics. The data acquisition will be generated at video rate, thereby allowing real-time visualization of the dynamics. In the R33 phase of the project, the same system design will be extended to multiple parallel detectors, allowing quantitative spectroscopy of oxygen saturation and total hemoglobin. The system will be used in existing MRI-NIR imaging systems for both brain and tumor imaging studies. The combined MRI-NIR imaging system will be calibrated in phantoms and software refined for rapid dynamic quantitative hemoglobin imaging with region-based targeting of the reconstruction. We hypothesize that dynamic imaging of total hemoglobin and oxygen saturation changes will provide fundamentally new information about tissue function and metabolic response to stimulus. The contrasts to be examined in rodent brain will be functional activation studies, as well as imaging of tumors will include (i) intrinsic blood pulsatile flow (ii) pulsatile flow and kinetics of indocyanine green, (iii) dynamic changes in oxygen saturation in response to long term and transient changes of inspired oxygen concentration, and (iv) glucose induced metabolic changes in oxygen consumption from the Crabtree effect.
| Li, Zhiqiu; Krishnaswamy, Venkataramanan; Jiang, Shudong et al. (2010) Rapid magnetic resonance-guided near-infrared mapping to image pulsatile hemoglobin in the breast. Opt Lett 35:3964-6 |
| Li, Zhiqiu; Krishnaswamy, Venkataramanan; Davis, Scott C et al. (2009) Video-rate near infrared tomography to image pulsatile absorption properties in thick tissue. Opt Express 17:12043-56 |
| Darling, A L; Yalavarthy, P K; Doyley, M M et al. (2007) Interstitial fluid pressure in soft tissue as a result of an externally applied contact pressure. Phys Med Biol 52:4121-36 |
