Tan S, Roth B J, Wikswo J P
Department of Physics and Astronomy, Vanderbilt University, Nashville, TN 37235.
Electroencephalogr Clin Neurophysiol. 1990 Jul;76(1):73-85. doi: 10.1016/0013-4694(90)90059-s.
We extend our theoretical model based on spatial filtering to determine the ability of a SQUID magnetometer to resolve 2 localized current sources in the brain. We find that in order to resolve 2 separated but coaxial cortical sources, the source-to-pickup coil distance must be comparable to the distance between the 2 sources. The size of the current sources affects the resolving power of a magnetometer, but an anisotropy of 3 in the cortical tissue does not produce a significant effect. The model also provides a solution of the inverse calculation, either to reconstruct the original current source distribution from the measured magnetic field, or to continue the field at the magnetometer inward towards the sources. Both the inverse and inward calculations are limited by the fact that the inverse filter function serves as a high-pass filter, which leads to instabilities at high spatial frequencies, particularly in the presence of noise. The instabilities can be minimized by choosing an appropriate window to attenuate the noise, but this in turn reduces the spatial resolution.
我们扩展了基于空间滤波的理论模型，以确定超导量子干涉仪磁力仪分辨大脑中两个局部电流源的能力。我们发现，为了分辨两个分离但同轴的皮质源，源到拾取线圈的距离必须与两个源之间的距离相当。电流源的大小会影响磁力仪的分辨能力，但皮质组织中3的各向异性不会产生显著影响。该模型还提供了一种逆计算解决方案，既可以从测量的磁场重建原始电流源分布，也可以将磁力仪处的场向内延续到源处。逆计算和向内计算都受到逆滤波函数作为高通滤波器这一事实的限制，这会导致高空间频率下的不稳定性，特别是在存在噪声的情况下。通过选择合适的窗口来衰减噪声，可以将不稳定性降至最低，但这反过来又会降低空间分辨率。
