Biomedical Engineering Reference
In-Depth Information
possible (900-1000 nm instead of 800 nm) even though such
wavelength does not necessarily give the best two-photon absorp-
tion cross-section. (i) The incident light scattering length also
decreases with the age of brain tissue. We found that it is divided
by two in the adult brain (3-month old) in comparison to 2-week
old brains. A good compromise is to use young adults in which
most neurophysiological processes are matured and in which the
scattering length has intermediate values. (ii) Imaging at large
depth raises a problem of fluorescence background: the high inci-
dent light intensity required to excite the chromophore in depth
becomes sufficient to excite chromophores at the surface or just
below, even if the beam is not focussed. As a result, the autoflu-
orescence or the fluorescence generated in cells or vessels on the
beam path may generate a background that decreases the signal-
to-noise ratio of the signal of interest. Finally, the best is to choose
the incident light experimentally i.e. to vary and choose the inci-
dent beam wavelength that gives the best signal-to-noise ratio
at the depth of interest. (iii) Imaging at large depth can also be
improved by redistributing the laser power with a regenerative
amplifier: the average power is kept constant while the laser pulse
energy is increased and the repetition rate decreased
(53, 54)
.
Note that the photodamage caused by such pulses in vivo still
need to be investigated. (iv) An additional problem worth men-
tioning concerns light aberrations resulting from refraction index
changes at the surface and in the tissue. These aberrations reduce
the resolution and efficiency of the incident light. They can par-
tially be corrected with adaptive optics
(55)
but it has not been
attempted in the mammalian brain in vivo.
3.2. Collection of
Fluorescence Light
If it is acknowledged that high numerical aperture objectives are
best for fluorescence collection, it is necessary that the angular
acceptance of the microscope detection system matches that of
the objective. Practically, the detection system must be as close as
possible to the objective back aperture. The gain in fluorescence
collection is then particularly interesting at large depth
(52)
.
4. Capillary
Distribution in the
Superficial
Olfactory Bulb
Layers
As shown in the cortex
(45)
, an I.V. injection of fluorescein- or
texas red dextran easily reveals the vascular architecture of the
dorsal OB. In order to classify the vessel types according to the
different OB superficial layers, we labeled glomeruli by intranasal
injections of Oregon green- or Texas red dextrans (PM 10,000)
1-5 days prior to microcirculation experiments
(28)
. Note that
fluorescence from nasal and vessel dyes were acquired on two
“green” and “red” channels (separated using a dichroic mirror
