Yoko Hoshi: “Development of time-domain diffuse optical tomography based on a radiative transfer equation and diffusion approximation hybrid”

July 18, 2014  17:00 〜 18:30

3F seminar room, Icho-kaikan

Yoko Hoshi

Integrated Neuroscience Research Project, Tokyo Metropolitan Institute of Medical Science


Diffuse optical tomography (DOT) has great potential for the quantitative detection of focal changes in cerebral hemoglobin (Hb). DOT can be performed with continuous wave (CW), time domain, and frequency domain spectroscopy instruments. Recently, high-density CW DOT has been developed and improved brain specificity compared with conventional optical topography. This CW DOT, which is based on linear-single step image reconstruction, provides only qualitative images of measured changes. Although this method is enough for functional neuroimaging, quantitative images of steady-state cerebral Hb are further useful for diagnostic neuroimaging. Thus, we have been developing time-domain DOT based on a non-linear iterative reconstruction scheme.

It has been widely accepted that the radiative transfer equation (RTE) accurately describes photon propagation in biological tissue, while because of its high computation load the diffusion equation (DE) is often used as a forward model. However, the DE is invalid in low-scattering and/or highly absorbing regions and the vicinity of sources, which reduces image quality.

One of the most promising alternative approaches is a hybrid model based on the RTE and DE, which has been proposed in frequency domain. Extending the concept of the hybrid model in the steady state to the time domain, we are developing a space-time hybrid model in random media under refractive-index mismatching. In this seminar, first I will briefly summarize near-infrared spectroscopy (NIRS), and then I will talk about time-domain DOT.