Home » Posts tagged 'Computational Neuroscience'
Tag Archives: Computational Neuroscience
Su-004-Data-informed Selective Channels DOT Prior for EEG Reconstruction Using Two Wavelengths
Depth-sensitivity of diffuse optical tomography (DOT) can be used to enhance localizing neural activity from electroencephalography (EEG). Yet, the choice of which channels to use in reconstruction is often ad hoc. We recently have shown that DOT-enhanced EEG reconstruction may be improved using a data-informed method to select optimal channels for DOT reconstruction. Here we expand to two wavelengths to reconstruct hemoglobin changes and compare reconstruction of DOT priors using different DOT channels. We also study the effects of using optimized DOT priors on spatial resolution of EEG reconstruction by analyzing the full width half maximum of DOT-enhanced EEG reconstructions. Continue reading
Su-020-Towards Semantic Encoding of Visual Content in Movies via High-Density Diffuse Optical Tomography
Visual semantic maps are related to language processing and cover extensive cortical areas. This study aims to develop very high-density Diffuse Optical Tomography (VHD-DOT) for visual semantic mapping in a naturalistic setting. Here, we demonstrate that VHD-DOT has sufficient signal power (high explained variance) for applying a visual semantic voxelwise encoding model. Continue reading
Su-068-Portable, High-Density, Whole-Head Diffuse Optical Tomography in Highly Naturalistic Settings
Though noninvasive neuroimaging has been working towards technologies that are high resolution, robust to motion, and compatible with naturalistic settings, there has remained a tradeoff between resolution and portability. Our new, wearable, high-density (WHD), whole-head DOT system offers the high spatial resolution of conventional DOT, portability, and robustness to motion. We demonstrate the system in vivo in laboratory conditions and naturalistic tasks, including walking, playing piano, and social interaction. Even in unconstrained, natural, non-laboratory environments with freely-moving participants, we detect brain activations consistent with known brain networks. Continue reading
Su-086-Unveiling spectral signatures of HbO oscillations in different sleep states using all-night EEG/fNIRS in healthy subjects
We propose whole-night EEG/fNIRS, targeting bilateral frontoparietal regions, to investigate cortical oscillations of oxyhemoglobin (HbO) during sleep. Five healthy subjects underwent whole night EEG/fNIRS. We applied Morse analytical wavelet to characterize the oscillatory modes of oxyhemoglobin in different sleep stages, while removing the aperiodic 1/f component. We found specific oxyhemoglobin oscillation signatures around 0.005 and 0.01Hz during tonic and phasic Rapid-Eye-Movement sleep respectively, 0.03-0.06Hz during N2 sleep, and ~0.2Hz during N3 sleep. We are reporting for the first-time sleep-state-specific spectral signatures of HbO oscillations, as a promising approach to characterize physiological and pathological sleep. Continue reading