Table 1.

Resolution metrics.

Dipole localization error	Spatial dispersion	Overall amplitude
$\begin{array}{l}\text{D}L{E}_i=\sqrt{{\left(x_{\text{p}}-x_{\text{i}}\right)}^2}\hfill \\ x_{\text{p}}:\text{peak coordinate}\hfill \\ x_{\text{i}}:\text{true coordinate}\hfill \end{array}$	$\begin{array}{l}{\text{SD}}_i=\sqrt{\frac{\sum _j{d}_{\mathit{ij}}{F}_{\mathit{ij}}^2}{\sum _j{F}_{\mathit{ij}}^2}}\hfill \\ d_{\mathit{ij}}:\text{distance betweenlocations }j\text{ and }i\hfill \end{array}$	$A_i=\sum _j\left|F_{\mathit{ij}}\right|$ (normalized to maximum for each subject)

Formulas for the resolution metrics used in our simulations. F represents the distributions of point-spread functions (PSFs) and cross-talk functions (CTFs), respectively. | | indicates the absolute value. PSFs are the columns of the resolution matrix, representing the estimated source distribution for a single point source. CTFs are rows of the resolution matrix, describing the sensitivity of an amplitude estimator for a single point source to all other sources.