Table 1.

Overview of the spatiotemporal resolution, specificity, observed patterns, and considerations for all of the discussed imaging modalities.

	Electrode arrays	Optical imaging of fluorescence and intrinsic signals	fMRI	EEG
Number of neurons	One to hundreds	Tens to hundreds	n/a	Millions, cortical surface
		n/a
Spatial resolution	200 to $400\mu \mathrm{m}$	50 to $100\mu \mathrm{m}$	400 to $500\mu \mathrm{m}$	10 to 20 mm
		$\sim 100\mu \mathrm{m}$
Time resolution	Millisecond	Millisecond (depends on kinetics of sensor)	Second	Millisecond to hundreds of milliseconds
		Second
Cell-type specificity	Limited	High (depends on the promoter)	Very low signal from hemodynamic response to neural activity.	Low signal predominantly from EPSPs and IPSPs in cortical pyramidal apical dendrites
		n/a
Patterns found	(1) Slow oscillation generated around layer 5 (in vitro ferret occipital cortex, in vivo rats and cats) 6,7,8,9,10	(1) Global plane waves traveling anterior to posterior (mice)13,14	(1) Propagating activity, such as QPPs (rats and humans)16,17,18–22	(1) Transient spatial configurations, such as EEG and microstates (humans)31,26,27,28,29,30
	(2) Brief spontaneous depolarizations, localized to an area of a barrel column (rats and mice)11	(2) Standing waves that have no net movement (mice)15	(2) Transient spatial configurations, such as CAPs (rats and humans)23,16,24,25
	(3) Propagating waves that traveled horizontally across dorsal cortex (rats and mice)11,12	(3) Local source, sink, and saddle points (mice)15
Pitfalls	(1) Large artifacts	(1) Problems of light delivery	(1) Low temporal resolution	(1) Low spatial resolution
	(2) Tissue damage	(2) Dependent on efficiency of viral transfection	(2) No cell type specificity	(2) Signal primarily from postsynaptic potentials of apical dendrites
		(3) Potential toxicity of opsins	(3) Low SNR	(3) Superficial signal, no localization at greater cortical depths
		(4) Unspecific effects