Image Reconstruction Strategy Of Continuous Wave Diffuse Optical Tomography For The Slab Sample Towards Near-Infrared Functional Brain Imaging

Near-infrared functional brain imaging can employ safe levels of optical radiation in the wavelength region of 650-900 nm and primarily interrogate the changes of the absorption coefficient in the grey matter induced by brain activation and potentially achieve spatial resolution of 10 mm in the human brain close to the skull. Recently, the popular scheme is optical topography(OT) based on the modified Larmbert-Beer law(MLBL), but the disadvantages of MLBL-OT imaging are low-precision and low-resolution. Since MLBL simply assumes the uniform distribution of optical parameters between the source site and the detector site, and does not effectively reflect the scattering effect of tissue. Due to the greater spatial resolution and quantitativeness, the near infrared brain imaging system based on diffuse optical tomography(DOT) gradually obtains more attention. Therefore, new reconstruction algorithms are investigated for the continuous wave(CW) DOT scheme to improve the quality of the reconstructed image with the slab sample.Firstly, the semi-3D image reconstruction algorithm is proposed to alleviate the underdetermined problem, according to the anatomic structure of head and the diffusive nature of near infrared light propagation in the brain. Namely, the dense mesh of 3D finite element method is used to improve the calculation accuracy in forward problem, and the sparse mesh of 2D image reconstruction algorithm is used in inverse problem by assuming that optical properties in the grey matter is invariable along the depth. The numerical and phantom validations results show that the 2D images acquire the higher spatial resolution and quantitativeness by using the semi-3D reconstruction algorithm and the reconstruction speed by using the semi-3D reconstruction algorithm is faster than the 3D reconstruction algorithm.Secondly, with the development of technology, multimodal imaging methods gradually become a new trend in the study of brain function, which can provide structural-prior information to with the DOT reconstruction. However, multimodal imaging methods increase the cost of instruments and the difficulties of the registration and fusion. MLBL-OT-guided DOT scheme(OT-DOT) is proposed to solve this problem with a ?functional priori‘ obtained by MLBL-OT images. Namely, a template matrix is introduced to the traditional Jacobian matrix of CW-DOT, which can be deduced from the functional image of MLBL-OT. The numerical and phantom validations are performed with this method, and the reconstruction results indicate that the proposed method can effectively alleviate the ill-posedness of the inverse problem and improve the quality of the reconstructed X-Y and X-Z sectional images of the common two-layered slab sample.Thirdly, the accuracy of the background optical properties has a considerable effect on the quality of reconstructed images in near-infrared functional brain imaging based on CW-DOT. But the background optical properties of the two-layered slab sample are derived from in vitro human brain tissues, which are different from those in living tissues. Region stepwise reconstruction method(RSRM) is proposed for reconstructing the background absorption and reduced scattering coefficients of the two-layered slab sample. Namely, the conventional measurement data are divided into two groups and employed to reconstruct the top and bottom background optical properties respectively according to the relation between the thickness of the top layer and source-detector-separation. The numerical simulation results demonstrate that the proposed method can reconstruct the background optical properties of two-layered slab sample effectively.Finally, we complete the experimental measurement of the phantom with a home-made CW-DOT system, and validate the feasibility of our proposed algorithm.