Biomedical Engineering Reference
In-Depth Information
Parameter Estimate
Data
600
600
500
500
400
400
300
300
200
200
100
100
0
5
10
15
20
0
5
10
15
20
(a)
Saccade Magnitude (Degrees)
(b)
Saccade Magnitude (Degrees)
Duration vs. Saccade Magnitude
Latent Period vs. Saccade Magnitude
100
350
300
80
250
200
60
150
y = 3.1102x + 26.546
100
40
y = 2.291x + 171.84
50
y = -0.29x + 43.929
20
0
0
5
10
15
20
0
5
10
15
20
(c)
Saccade Magnitude (Degrees)
(d)
Saccade Magnitude (Degrees)
FIGURE 13.53
Main sequence diagram for all 127 saccades from three human subjects. (a) Peak velocity versus
saccade magnitude from the model estimates, with regression fit
:
y
pv
¼
e
0:2y
ss
390 1
(b) Peak velocity versus
:
y
pv
¼
e
0:2y
ss
saccade magnitude from the data, with regression fit
(c) Duration versus saccade magnitude
based on the data. (d) Latent period versus saccade magnitude based on the data. Note that the parameter
estimation program did not update the duration or the latent period, so a single graph for each is drawn.
401 1
end of off-saccades in monkeys in other studies (for examples, see [42]and[46]). It was noted
earlier that saccades with dynamic overshoots or glissades do not occur with the same
frequency in the monkey as in humans, and that they are absent from our monkey data.
The theory presented here is that, at least in humans, the antagonist PIRB causes a
reverse peak velocity during dynamic overshoots or glissades in humans. The model pre-
dictions accurately match the velocity data for the entire saccade, including saccades with
dynamic or glissadic overshoot. We were unable to generate saccades with postsaccade
behavior based on just the timing of the antagonist step, but we needed the PIRB to gener-
ate saccades with dynamic or glissadic overshoot.
Figures 13.55 and 13.56 summarize the characteristics of the 127 saccades collected from
the three human subjects. The number of saccades with a glissade is larger than the number
of normal saccades or those with a dynamic overshoot. Additionally, the incidence of
dynamic overshoot decreases as saccade size increases. As shown, saccades with a dynamic
overshoot typically have larger rebound burst magnitude than those with a glissade or with
