Email: Jenni.Heino@hut.fi
CompanyName: Helsinki University of Technol
Country: Finland
Abstract Title:
Simultaneous estimation of optical anisotropy and absorption
in medical optical tomography
Authors: Jenni Heino 1, Erkki Somersalo 2, Simon Arridge 3
1 Laboratory of biomedical engineering, Helsinki University of Technology,
P.O.Box 2200, 02015 HUT, Finland
2 Laboratory of Mathematics, Helsinki University of Technology,
P.O.Box 1100, 02015 HUT, Finland
3 Department of Computer Science, University College London,
Gower Street, London WC1E 6BT, UK
Abstract
Optical tomography is a relatively new, non-invasive modality for
medical applications such as functional imaging of the brain
or breast cancer detection. Techniques using near-infrared light
have some advantages over some of the existing modalities.
Near-infrared light is non-ionizing and thus harmless to the
patient, enabling long term monitoring. Also, the instrumentation
can be made relatively light and inexpensive.
At near-infrared wavelengths, most human tissues are highly
scattering. The propagation of light is often modelled using the
Radiative Transfer Equation (RTE) and its approximations. The simplest
approximation often used to model the forward problem in optical
tomography is the Diffusion Approximation (DA), which is obtained
by first order spherical harmonic expansion of the RTE, and making
the additional assumption that the scattering probability depends only
on the relative angle of incident and scattered radiation, not on the
absolute direction. Such a medium is often referred to as isotropic.
However, for several human tissues, such as the white matter of the
brain, muscles, or skin, the assumption of isotropic medium may not
hold. It is known, e.g., that the fibres in the white matter of the
brain have direction dependent properties for the diffusion of
water. Here, we present one approach to model the
anisotropic effects of light propagation and some examples of
simultaneous reconstruction of optical absorption and anisotropy
model parameters.