Physiological Modeling - Introduction to Biomedical Engineering - page 890

0.45

0.4

0.35

0.3

0.25 F ant

0.2

N ant

0.15

0.1

0.05

0
0.2

0.25 0.3 0.35 0.4

(e)

Time (s)

FIGURE 13.49, cont'd

13.8.3 Monkey Data and Results
Data 1 were collected from a rhesus monkey that executed a total of 27 saccades in our
data set for 4 ,8 ,16 , and 20 target movements. Neuron data were recorded from the long
lead burst neuron (5 saccades), excitatory burst neuron (17 saccades), and the agonist burst-
tonic neuron (5 saccades) (the neuron types are described in Section 13.9). The firing of
the burst-tonic neuron is similar to the motoneuron that drives the agonist muscle during
a saccade. Figure 13.50 shows the estimation results for three saccades (4 ,8 , and 15 ).
Figure 13.51 shows the estimated neural inputs and active-state tensions that generate
the saccades shown in Figure 13.50. Also shown are the firing rates recorded from a single
burst-tonic cell in a rhesus monkey for these saccades, scaled to match the height of

N ag .
The shapes of the model's neural inputs approximate the burst-tonic data during the pulse
and slide very closely. The estimated agonist neural input clearly has a similar shape to

the firing rate data. It should be noted that the firing activity in the data comes from a single
burst-tonic neuron. The neural input to the oculomotor plant is actually due to the firing of
more than 1,000 motoneurons.

N ag

13.8.4 Human Data and Results
Data 2 were collected from three human subjects executing 127 saccades, many with
dynamic overshoots or glissades. Figure 13.52 shows representative model estimates of sac-
cades generated with a dynamic overshoot, a glissadic overshoot, and normal characteristics.

1 Details of the experiment and training are reported in Sparks et al. [44]. Data provided personally by
Dr. David Sparks.
2 Details of the experiment are reported in Enderle and Wolfe [17].