Journal: Journal of Neurophysiology
Article Title:
doi: 10.1152/jn.00033.2020
Figure Lengend Snippet: Overall accuracy of the visually evoked potential (VEP) template-matching algorithm. A: distribution of absolute errors in estimating the depth of the recording sites using our method, as obtained by computing the absolute distance between predicted and measured depth for every site on every probe (the dashed line shows the mean of the distribution). B: average accuracy (across the 18 recording sessions) in predicting the laminar location of the recording sites ± SE, as measured for 3 different levels of distinction between the cortical layers, from fine to binary (see key). The dark gray bars report the accuracy achieved by our automated, template-matching method, and the light gray bars display the accuracy attained by visually inspecting the current source-density (CSD) patterns and VEP profiles. The dashed lines show the accuracy that would be obtained by guessing (i.e., the chance levels). C: the overall recall (black) and precision (gray) accuracies yielded by our method in predicting the laminar location of the recording sites are reported as a function of the measured (for the recall) or predicted (for the precision) cortical layer of the sites (see main text for a definition of recall and precision). The white number on a black bar indicates the number of recording sites in the corresponding layer (or group of layers), as measured through histology (this corresponds to the denominator in the formula to compute the recall, i.e., true positives plus false negatives). The white number on a gray bar indicates the number of sites that our method assigned to the corresponding layer (this corresponds to the denominator in the formula to compute the precision, i.e., true positives plus false positives).
Article Snippet: Through the years, a technique known as “current source-density” (CSD) analysis has been widely exploited to fulfill this goal in different animal models, from monkeys ( Hansen et al. 2011 ; Schroeder et al. 1998 , 2001 ; Takeuchi et al. 2011 ) to rodents ( Dadarlat and Stryker 2017 ; Deliano et al. 2020 ; Happel et al. 2010 ; Headley and Weinberger 2015 ; Hoy and Niell 2015 ; Li et al. 2018 ; Niell and Stryker 2008 ; Sieben et al. 2013 ).
Techniques: