A Simulation Study of Optode Placement in Functional Near Infrared Spectroscopy

• Main Authors: Yu-Ming Lai, 賴鈺銘
• Other Authors: 宋孔彬
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/7388p3

碩士 === 國立臺灣大學 === 生醫電子與資訊學研究所 === 104 === Functional near-infrared spectroscopy (fNIR,) technic has developed rapidly recent year. fNIR can use in open-environment which no other neuroimaging can use in. Hardware and algorithm is progressing in fNIR, but optode placement, which also cause significant influence in reconstruction, is researched less. This research try to solve the problem how source and detector placement in fNIR can improve the sensitivity. First, using well-known tessellation in geometry to derivate the geometry of optode placement. Then, let one of the channel in optode placement become 2.9 cm, which is the best source-detector separation (SDS) in brain model. After that, we provide 4 characteristics─ channels per cm2, optodes controlled per cm2, channels per optode and kinds of SDSs─ to analysis the performance of optode placement. Following the characteristics, we development a progress to select the placement. It is the correctness of the progress that is proved by AtlasViewer simulation program. To sum up, in the density of optode opinion, hexagonal placement with SDS 2.9 cm, source separation(SS) 5.8 cm is the best in low density. Square placement with SDS 2.05 cm, SS 4.5 cm is the best in medium density. In high density, square placement with SDS 1.45 cm, SS 324 cm is the best. Square placement with SDS 1.3 cm, SS 1.83 cm is the best in super high density. By summarizing the process and the conclusion, we can follow the rule in optode placement and modify by user preference. As a result, the signal in fNIR is stronger, and the performance of reconstruction is better. In future, an automatic analysis program with the progress is needing to development new combined placement.