Biomedical Engineering Reference
In-Depth Information
filled with 1.2% agarose (Agarose-HGS, Nacalai tesque, Japan,
gel strength 1.5%) containing dexamethasone and antibiotics and
was sealed with a round glass cover slip and a silicon gasket. Rec-
tal temperature, ECG and expired CO 2 were continuously moni-
tored during both OCT experiments and surgery.
OCT Scan
The stimuli were the same as that used for OISI and consisted
of square-wave gratings (white = 8 cd/m2, black=0 cd/m2) hav-
ing a spatial frequency and moving at a velocity of 4 degrees/sec.
The stimulus set consisted of five patterns with control or blank
(mean luminance 4 cd/m 2 ), horizontal (0 ), vertical (90 )and
oblique gratings (45 , 135 ) and were presented in a random
order. All stimuli were generated with a VSG2/3 graphics video
board (Cambridge Research Systems, UK). The center of the
visual field was roughly estimated by projecting images of optic
discs onto a screen in front of the animal. The distance of the CRT
screen (200-300 mm) was adjusted to have the best focus of optic
discs and surrounding vessel patterns. A total of 40 trials were
obtained for each stimulus. In a single trial, data acquisition was
done for 8 s, during which time 16 x-z frames (128
×
100 pixels
corresponding to 1
1 mm) were obtained. The inter-stimulus
interval (ISI) was 5 s. A schematic is shown in Fig. 6.7D . Stim-
uli appeared with a delay of 2 s after the acquisition onset and
persisted for 2 s.
×
OCT Data Correction
and Analysis
First, the scans were corrected for any misalignment of the surface
position by a correlation-based procedure. For a single stimulus,
we obtained a total of 640 scans. To compensate for the small
variations in the surface position over different scans, we used
correlation analysis. Out of these 640 scans, we selected an arbi-
trary scan ( i ) with the reflectivity detected as a function of depth
being R s ( x,z,i ) at a lateral position x for a stimulus s. Next, we cal-
culated the correlation between the i th scan and rest of the scans
( j ) using the following equation:
EMBED Equation.DSMT4
−∞
R s ( x , z , i ) R s ( x , z
+
z , j ) dz
−∞
(
z ; i , j )
=
(6.1a)
R s ( x , z ; i ) dz
R s ( x , z
+
z ; j ) dz
This operation gives the position of the correlation peak that
corresponds to the amount of shift necessary for the j th profile to
be in alignment with the i th profile.
 
Search WWH ::




Custom Search