| Project/Area Number |
15300169
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Biomedical engineering/Biological material science
|
|---|
| Research Institution | Sophia University |
|---|
Principal Investigator |
NAKAYAMA Kiyoshi Sophia University, Faculty of Science & Technology, Professor, 理工学部, 教授 (00053653)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
OZEKI Takeshi Sophia University, Faculty of Science & Technology, Professor, 理工学部, 教授 (40245791)
FUJII Mamiko Sophia University, Faculty of Science & Technology, Lecturer, 理工学部, 講師 (20173396)
FUKUDA Keiko Tokyo Metropolitan College of Aeronautical Eng., Associate Professor, 電子工学科, 助教授 (70396266)
鈴木 彰文 上智大学, 理工学部, 助手 (00221317)
|
|---|
| Project Period (FY) |
2003 – 2005
|
| Project Status |
Completed (Fiscal Year 2005)
|
| Budget Amount *help |
¥9,900,000 (Direct Cost: ¥9,900,000)
Fiscal Year 2005: ¥3,900,000 (Direct Cost: ¥3,900,000)
Fiscal Year 2004: ¥4,300,000 (Direct Cost: ¥4,300,000)
Fiscal Year 2003: ¥1,700,000 (Direct Cost: ¥1,700,000)
|
|---|
| Keywords | Near Infrared / brain activity / diffuse optical imaging / 3D reconstruction / M-P generalized inverse / depth-adaptive regularization / 拡散反射型光CT / 逆問題 / 正則化パラメータ / 再構成画像ノイズ / 空間分解能 / Moore-Penrose / 単純正則化Moore-Penrose一般逆行列 |
|---|
| Research Abstract |
This project is a feasibility study on diffuse optical tomography which has 3-D spatial resolution using 2-dimensional CW source-detector arrays on the body surface, aiming at real time imaging of the brain activation minimally contaminated by skin circulation fluctuation. We formulated the forward problem based on the Rytov approximation for the small perturbation of the absorption coefficients in the diffusion equation approximation of the photon migration, and formulated the 3-D tomographic reconstruction using simple regularized Moore-Penrose pseudo inverse. We analyzed the dependencies of the reconstruction error and the spatial resolution on the regularization parameter, the target depth and the reconstruction voxel size. The study showed that the trade-off between the reconstruction error and the spatial resolution does not depend on the voxel size. The simple regularized inverse, however, was found to have strong depth dependencies of the reconstruction sensitivity so that the
… More
target brain signal in the depth would be easily masked by the unwanted skin circulation signal in the shallow range. We devised the depth-adaptive voxel size method and the depth-adaptive regularization, both of which were shown to be effective to overcome this serious problem. Based on these extensive simulation studies, we clarified trade-offs between spatial resolution, reconstruction noise, reconstruction region of interest and the uniformity of the reconstruction sensitivity etc. To assess the feasibility experimentally, we modified a existing commercial optical topograph, enabling digital data acquisition for the reconstruction, and checked the performance by phantom study. Finally, we designed a practical imaging system with the source-detector system of the existing topograph. The system would visualize the brain activity with the spatial resolution of 1cm in the 4cm X 4cm area of the cerebral cortex assumed at the 1cm depth from the skin surface with the 20 dB suppression of the unwanted skin circulation signal. Less
|
