Phantom Experiment-Based Validation for Time-Varying Acoustoelectric Brain Imaging With Non-Stationary Current Source Characteristics

Abstract

Based on the acoustoelectric (AE) effect, acoustoelectric brain imaging (ABI) is a potential neuroimaging method that can be used to map brain electrical activity with high spatiotemporal resolution. Brain electrical activity is non-stationary, with time-varying frequency and amplitude. However, the current ABI mainly focuses on the current source with stationary current source characteristics. To make a step towards developing ABI as a clinical imaging technique, time-varying ABI with non-stationary current source characteristics is evaluated and verified. A decoded algorithm based on wavelet transform was adopted for AE signal to extract non-stationary current sources characteristics. To test the performance of time-varying ABI, three phantom experiments were designed with a time-varying current source, including one source with time-varying frequency, one source with time-varying amplitude, and two sources of different amplitudes with time-varying frequency. For the experiment of source with time-varying frequency, the time-frequency diagrams of the decoded signal presented clear power enhancements at specific frequencies, which was consistent with the frequency change of the current source. For the experiment of source with time-varying amplitude, the correlation coefficient between the decoded signal and the current source was 0.90. In addition, the SNRs of the decoded signal of current source positions were significantly higher than that of non-current source positions. Finally, the positions of sources with time-varying characteristics could be accurately located at different moments. These results validate the feasibility of time-varying ABI with non-stationary current source characteristics, which make a key step toward its application in neuroimaging.