Design and evaluation of a GUI for signal and data analysis of mobile functional near-infrared spectroscopy systems

Thumbnail Image
Issue Date
Moradi, Mousa
Desai, Jaydip M.

Functional near-infrared spectroscopy (fNIRS) is an emerging neuro-imaging modality that can indicate cortical functionality with good temporal resolution (0.5-1 sec). Unlike fMRI, up to now, there is no standard method and a whole package that can perform signal processing and data analysis for fNIRS dataset. This thesis is focused on developing a Graphical User Interface (GUI) model in MATLAB that can perform both signal and data analysis for data collected by mobile fNIRS devices either in offline or online mode. It can remove motion artifacts using detrending methods and performs Band Pass Filtering using the low-and high-pass cutoff frequencies of 0.6 and 0.01 Hz, respectively and check the quality of signal in term of Signal to Noise Ratio (SNR). The filtered fNIRS signals were then sent to a General Linear Model (GLM) to estimate the linear model parameters. T-statistical values have been calculated and unrelated values have been further removed prior to obtaining results in terms of evoked related potentials, 2D and 3D head projections. In this thesis, signal processing and data analysis have assessed in the proposed GUI through the data of four participants [three males, age of 22, 24, 28 and, one female, age of 19] collected from two different fNIRS systems. Results for both systems verified when signal is de-trended using Discrete Cosine Transform (DCT) and it is filtered using a BPF order 4, cutoff frequencies of 0.01 to 0.6 Hz) the SNR value of 6.99 dB from ARTINIS system was achieved. Brain mapping results are shown with both hand motor imagery, the subjects consumed more hemoglobin in their both hemispheres. To localize each optode position, more accurately and complicated methods like 3D digitizer can be replaced with 10-20 standard system. With advantages such as hazard-free application, motion artifact robustness and the possibility to miniaturize hardware, fNIRS may enable new applications in the BCI context, in clinical tools and diagnosis and further contribute to new insights in brain research and neuroscience.

Table of Content
Thesis (M.S.)-- Wichita State University, College of Engineering, Dept. of Biomedical Engineering