Wojtkiewicz Stanislaw, Bejm Karolina, Liebert Adam
Nalecz Institute of Biocybernetics and Biomedical Engineering, Warsaw, Trojdena 4, 02-109, Poland.
Biomed Opt Express. 2022 Mar 4;13(4):1869-1887. doi: 10.1364/BOE.448038. eCollection 2022 Apr 1.
Here we show a method of the lock-in amplifying near-infrared signals originating within a human brain. It implies using two 90-degree rotated source-detector pairs fixed on a head surface. Both pairs have a joint sensitivity region located towards the brain. A direct application of the lock-in technique on both signals results in amplifying common frequency components, e.g. related to brain cortex stimulation and attenuating the rest, including all components not related to the stimulation: e.g. pulse, instrumental and biological noise or movement artefacts. This is a self-driven method as no prior assumptions are needed and the noise model is provided by the interfering signals themselves. We show the theory (classical modified Beer-Lambert law and diffuse optical tomography approaches), the algorithm implementation and tests on a finite element mathematical model and on healthy volunteers during visual cortex stimulation. The proposed hardware and algorithm complexity suit the entire spectrum of (continuous wave, frequency domain, time-resolved) near-infrared spectroscopy systems featuring real-time, direct, robust and low-noise brain activity registration tool. As such, this can be of special interest in optical brain computer interfaces and high reliability/stability monitors of tissue oxygenation.
在此,我们展示了一种用于锁定放大源自人类大脑内部的近红外信号的方法。这意味着使用固定在头部表面的两个相互成90度角旋转的源 - 探测器对。这两个对都有一个朝向大脑的联合敏感区域。将锁定技术直接应用于这两个信号会放大共同频率成分,例如与大脑皮层刺激相关的成分,并衰减其余成分,包括所有与刺激无关的成分:例如脉搏、仪器和生物噪声或运动伪影。这是一种自驱动方法,因为无需事先假设,并且噪声模型由干扰信号自身提供。我们展示了该理论(经典修正比尔 - 朗伯定律和扩散光学层析成像方法)、算法实现以及在有限元数学模型上和视觉皮层刺激期间对健康志愿者的测试。所提出的硬件和算法复杂度适合于(连续波、频域、时间分辨)近红外光谱系统的全谱,这些系统具有实时、直接、稳健且低噪声的脑活动记录工具。因此,这在光学脑机接口和组织氧合的高可靠性/稳定性监测器方面可能具有特殊意义。