Assessing Cognitive Performance during Dynamic Activities with Varying Insole Textures

Abstract

Assessing the data quality and overall presence of current excited signal is an important stage of screening and processing the data produced from Magnetic Resonance Electrical Impedance Tomography (MREIT) experiments. To do this, Signal to Noise (SNR) analysis and two methods of subtractive comparison were utilized. The data analyzed consisted of previously collected MREIT scans of 12 subjects’ brains with magnitude and phase data in the form of magnetic flux density (Bz). The data was collected using a standard MREIT protocol with electrode positions at the F3, F4, and Oz locations following the 10-20 international standard. To perform subtractive comparison within each subjects’ data, the scan in which current was passed from the F4 to the Oz electrodes was subtracted from the scan where current was passed between the F3 and Oz electrodes. The result of this subtraction produced assembled Bz scans that with current excitations similar in both magnitude and structure to the experimental scan where current was passed between F3 and F4 electrodes. Two methods of statistical analysis were used to validate this outcome, Structural Similarity Index Measures (SSIM) and mean subtractive comparison between the experimental and assembled F3-F4 scans. Both of these measures showed statistical significance while a linear regression between a subject’s SNR and SSIM scores did not yield significant results. The two main forms of comparison presented show strong promise as useful tools in assessing the consistency and quality of MREIT data sets.

Video