Strategies for improving depth-penetration of two-photon imaging in vivo

We investigate tissue and instrument parameters affecting the penetration depth in two-photon microscopy. We show that the temporal redistribution of the same average power into fewer pulses of higher peak energy by means of a regenerative amplifier results in an increase in excitation depth by approximately 2-3 scattering mean free paths. We then measure the excitation scattering mean free path in vitro, using rat brain slices, as a function of the excitation wavelength and tissue age. We find that young-animal tissue (< P18) is two-fold less scattering than adult tissue (P90). We quantify the fall-off of the collected fraction of generated fluorescence in a backward detection geometry, in vivo. At large depths, we observe that the collected fraction scales as the angular acceptance squared (related to the effective field-of-view) of the detection optics. Matching the angular acceptance of the detection optics to that of the objective (63X NA-0.90) results in a factor 3-4 of the collected fluorescence. The collection efficiency can be further increased (10X) by using an objective with large field-of-view and high numerical aperture (20X NA-0.95). These gains translate into approximately 120 micrometers additional depth penetration when working in the rat brain in vivo with a standard Ti:sapphire source.