Comparison of human population receptive field estimates between scanners and the effect of temporal filtering.
Comparison of human population receptive field estimates between scanners and the effect of temporal filtering.
复制标题
DOI：
10.12688/f1000research.20496.1
复制
发表时间：
2019-01-01
期刊：
影响因子：
--
通讯作者：
Schwarzkopf, D Samuel
中科院分区：
文献类型：
--
作者：
Morgan, Catherine;Schwarzkopf, D Samuel
背景：基于功能性磁共振成像（fMRI）的群体感受野（pRF）分析是一种越来越流行的用于绘制视野表征和估计人类视觉皮层体素空间选择性的方法。然而，大量的实验设置和处理方法使得比较研究之间的结果变得困难。方法：在这里，我们比较了在两个不同国家的1.5特斯拉和3特斯拉扫描仪上使用可比扫描参数获得的相同三个个体的pRF图。我们还测试了时间序列的低通滤波对pRF估计的影响。结果：正如预期的那样，3个特斯拉数据的信噪比优越；然而，关键的是，对pRF大小和皮质放大率的估计并没有显示出这些部位之间的任何系统性差异。不出所料，低通滤波提高了拟合优度，可能是通过去除高频噪声。然而，低通滤波后含有有意义的视网膜异位信号的体素数量并没有显著增加。重要的是，滤波还增加了早期视觉区域中pRF大小的估计，这可能会严重扭曲对空间调谐特性的解释。结论：因此，我们的结果表明，在不同场强的扫描仪之间，pRF估计值通常具有可比性，但应谨慎使用时间滤波。
Background: Population receptive field (pRF) analysis with functional magnetic resonance imaging (fMRI) is an increasingly popular method for mapping visual field representations and estimating the spatial selectivity of voxels in human visual cortex. However, the multitude of experimental setups and processing methods used makes comparisons of results between studies difficult. Methods: Here, we compared pRF maps acquired in the same three individuals using comparable scanning parameters on a 1.5 and a 3 Tesla scanner located in two different countries. We also tested the effect of low-pass filtering of the time series on pRF estimates. Results: As expected, the signal-to-noise ratio for the 3 Tesla data was superior; critically, however, estimates of pRF size and cortical magnification did not reveal any systematic differences between the sites. Unsurprisingly, low-pass filtering enhanced goodness-of-fit, presumably by removing high-frequency noise. However, there was no substantial increase in the number of voxels containing meaningful retinotopic signals after low-pass filtering. Importantly, filtering also increased estimates of pRF size in the early visual areas which could substantially skew interpretations of spatial tuning properties. Conclusion: Our results therefore suggest that pRF estimates are generally comparable between scanners of different field strengths, but temporal filtering should be used with caution.