9. Saccade abnormalities in Gilles de la Tourette syndrome

Human brain mapping with near infrared spectroscopy (NIRS) measurements poses two major problems that limit its effectiveness for localizing cortical activation. One is lower spatial resolution and the other is artifacts due to physiological noises most exemplified by non-cortical skin blood flow changes. In this study, we propose a novel solution for these problems by simultaneously implementing a high-density probe arrangement and independent component analysis (ICA)-based signal processing. To achieve high spatial information density, oxygenated hemoglobin signals during a fast-finger-tapping (4 Hz) task were measured with double-density optical topography (DDOT), which consists of data from two alternately embedded single-density optical topographies (SDOT). Then, the measured signals were subjected to ICA to extract the signal components representing cortical activation. DDOT data can be decomposed into two different sets of SDOT data. Comparisons between SDOT and DDOT data revealed that DDOT portrayed both the spatially focused activation and the temporal structure relevant to hemodynamic response, whereas the SDOT data could only portray one or the other. The spatial resolution of DDOT might appear excessive, but it does indeed enable more efficient source separation than SDOT, leading to more robust spatial and temporal analyzes.