Song Xizi, Wang Tong, Su Mengyue, Chen Xinrui, Liu Xiuyun, Ming Dong
Academy of Medical Engineering and Translational Medicine, Tianjin University, Tianjin, China.
Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China.
Front Physiol. 2022 Dec 8;13:1054103. doi: 10.3389/fphys.2022.1054103. eCollection 2022.
Acousticelectric brain imaging (ABI), which is based on the acoustoelectric (AE) effect, is a potential brain function imaging method for mapping brain electrical activity with high temporal and spatial resolution. To further enhance the quality of the decoded signal and the resolution of the ABI, the decoding accuracy of the AE signal is essential. An adaptive decoding algorithm based on Fourier fitting (aDAF) is suggested to increase the AE signal decoding precision. The envelope of the AE signal is first split into a number of harmonics by Fourier fitting in the suggested aDAF. The least square method is then utilized to adaptively select the greatest harmonic component. Several phantom experiments are implemented to assess the performance of the aDAF, including 1-source with various frequencies, multiple-source with various frequencies and amplitudes, and multiple-source with various distributions. Imaging resolution and decoded signal quality are quantitatively evaluated. According to the results of the decoding experiments, the decoded signal amplitude accuracy has risen by 11.39% when compared to the decoding algorithm with envelope (DAE). The correlation coefficient between the source signal and the decoded timing signal of aDAF is, on average, 34.76% better than it was for DAE. Finally, the results of the imaging experiment show that aDAF has superior imaging quality than DAE, with signal-to noise ratio (SNR) improved by 23.32% and spatial resolution increased by 50%. According to the experiments, the proposed aDAF increased AE signal decoding accuracy, which is vital for future research and applications related to ABI.
基于声电（AE）效应的声电脑成像（ABI）是一种具有潜力的脑功能成像方法，可用于以高时间和空间分辨率绘制脑电活动。为了进一步提高解码信号的质量和ABI的分辨率，AE信号的解码精度至关重要。提出了一种基于傅里叶拟合的自适应解码算法（aDAF）以提高AE信号的解码精度。在所提出的aDAF中，首先通过傅里叶拟合将AE信号的包络分解为多个谐波。然后利用最小二乘法自适应地选择最大的谐波分量。进行了几个虚拟实验来评估aDAF的性能，包括具有不同频率的单源、具有不同频率和幅度的多源以及具有不同分布的多源。对成像分辨率和解码信号质量进行了定量评估。根据解码实验的结果，与带包络的解码算法（DAE）相比，解码信号幅度精度提高了11.39%。aDAF的源信号与解码定时信号之间的相关系数平均比DAE提高了34.76%。最后，成像实验结果表明，aDAF具有比DAE更好的成像质量，信噪比（SNR）提高了23.32%，空间分辨率提高了50%。根据实验，所提出的aDAF提高了AE信号的解码精度，这对于未来与ABI相关的研究和应用至关重要。
