Information Technology Reference
In-Depth Information
qualitative characteristics as power distribution ratios or correlation coefficients.
So far, the quantitative analysis, e.g., estimation of respiratory output using opti-
cal imaging data, has not been reported. In the present study, we aimed to develop
a new method to objectively and quantitatively evaluate voltage-imaging data. For
this purpose, we intended to construct a parametric model to analyze optical time
series data.
We have chosen the respiratory neuronal network in the brainstem as a represen-
tative system to test our method. Because the respiratory neuronal network in the
brainstem consists of functionally and anatomically distinct neuronal groups and
also forms motor nerve activity as the neural output, the respiratory neuronal net-
work is ideal as a model system for our analysis. The essential current knowledge
on the respiratory neuronal network in the brainstem is as follows. The respiratory
rhythm and motor patterns are generated by neuronal aggregates that are distributed
bilaterally in a columnar manner in the ventrolateral and dorsolateral reticular for-
mation (for review see [24, 7, 6, 8]). In neonatal animals, two respiratory related
ventrolateral medullary regions, the parafacial respiratory group (pFRG) [23] and
the pre-Botzinger complex (pre-BotC) [26], have been identified as putative RRGs.
However, the detailed function and anatomy of these RRGs have not been clarified.
12.2 Methods
12.2.1 Recording of Optical Signals and Preprocessing
We recorded respiratory neuronal activities in the brainstem by voltage imaging in
isolated brainstem-spinal cord preparations. For principles and general techniques
of voltage imaging, refer to the reviews by Cohen and Salzberg [4], Kamino [15],
Ebner and Chen [5], and Baker et al. [1]. Briefly, preparations were made of neona-
tal Sprague-Dawley rats (
n
19, 0-1 day old) under deep anesthesia as described
elsewhere [27, 20, 18, 19, 22]. Experimental protocols were approved by the Animal
Research Committee of Hyogo College of Medicine. Preparations were stained with
a voltage-sensitive dye (di-2-ANEPEQ) [19,22]. Inspiratory burst activity was mon-
itored from C4VR using a glass suction electrode. Activity of respiratory neurons in
the ventral medulla was analyzed using an optical recording system (MiCAM Ul-
tima, BrainVision, Tokyo). Preparations were illuminated with a tungsten-halogen
lamp (150 W) through a band-pass excitation filter (
=
λ
=
480-550 nm). Epifluores-
cence through a long-pass barrier filter
590 nm) was detected with a CMOS
sensor array. Magnification of the microscope was adjusted to 2.8
(
λ
>
×
3.3
×
depending
×
−
×
on the size of the brainstem. One pixel corresponded to 30
30
35
35
μ
m, and
5mm
2
. A total of 256 frames, 50
frames/s, were recorded starting at 1.28 s before the onset of C4VR activity.
As shown in Fig. 12.2b, c, raw optical signals had poor signal-to-noise ratios
and respiratory related signals were not obvious. The correlation coefficient values
×
−
.
×
.
the image sensor covered a total of 3
3
3
5
3
