We now turn to the question of the role of the computer in data acquisition, reduction and analysis. According to our knowledge, the properties of the nervous system that we are trying.
ANTERIOR
HORIZONTAL CROSS-SECTION
DORSAL
VERTICAL CROSS-SECTION
In vertebrates, movement in the preferred direction simply causes a higher discharge rate than movement in the null direction [2]. Collett and Blest [15] also report finding oppositely directed units close together in the optic lobe of the moth.
A summary of the Ila-in class response to movement as a function of pattern position is given in the contour map of Figure 3. The contour map shows the response of the Ila-in unit (firing rate) to a small pattern lying along a trajectory of the great circle [ 8] As a consequence, such cases of multiple unit registrations are usually presented in the form of.
Experiments with the cat's cochlear nucleus showed that the interaction. For the majority of the studies described herein, signals were sent directly to the computer during the experiment. Signals are playing from the FM tape recorder, part of which can be seen at the bottom of the photo.
The sampling rate for the waveform is independent of the clock rate for the TOEs. With the exception of the system of application programs, details of the other systems appear elsewhere and, with. Another consequence of the same decision is that the nature of the interaction between.. the user and his data is almost entirely a function of the application process.
MCG+
At the end of the run, the result would be displayed as in Figure 4. 13 (a), which shows the complete record. The user can now view selected parts of the record (Figure 4. 13 (b)), check the validity of the answers, review . details, compare both responses, etc. After the PST histogram is estimated, new stimulus conditions can be set, possibly dependent on the estimate, and the process repeated.
Here, peaks are represented by vertical lines (bars) between the maximum and minimum points of the complete. After selecting the characteristic, the user must specify ranges of allowable values for the characteristic for each source. As mentioned in Chapter I, one of the objectives of this work was to develop a system of application processes that would be.
We conclude with a diagrammatic summary of the data collection and analysis system as shown in Figure 4.
CHAPTER V
WW enc
LAG (MSEC) (b)
Because of the one-to-one relationship between spike trains, it does not seem likely that the data is from two. After the initial placement of the electrode, the signal being recorded was examined on the oscilloscope. The similarity of waveforms in the brain electrode made them inseparable from waveform methods.
The difference in size was due to the fact that the stimulus was located in the right eye field. With inward movement in the left eye field, the relative sizes of the correlogram peaks would be reversed. Synchronization of the lobe peaks with some brain peaks indicated that the latter originated from Illa and lllb units.
9(a) and (b) show the appearance of spikes before and after height sorting for the two stimulus conditions.
ELECTRODE
In the figure, the expected number of coincidences for uncorrelated records is marked by the letter "E". Significant features in the correlogation should thus be several bins wide, and relatively independent of small changes in resolution [67]. The valley on the right indicates that the unity in the left lobe has an inhibitory effect on the unity in the right lobe, and vice versa.
The valley to the right (with a delay of about 15 msec} indicates that the unit in the left lobe has an inhibitory influence on the unit in the right lobe, and vice versa. More precisely, the right valley indicates a decreased probability of a spike from the unit in the right lap given the occurrence of a spike in the second unit The lower correlogram is for unshifted data, in the upper one a spike train has been shifted in time by five seconds.
While the inhibitory valleys are clear in the lower, the upper correlogram is essentially flat.
MUTURL INHIBITION NET
The model was positioned so that it was in the field of only one of the Ila-in units. While the right unit responds strongly to internal movement, the speed of the left unit decreases. Note that there was no change in mean rate for either unit as the direction of the pattern was reversed.
Second, the absence of a rate change for the left unit (the right eye is covered) showed that for the noncontrol case it was not directly affected by the stimulus, but indirectly by the response of the right unit. Two movement patterns were used, each only in the field of the corresponding Ha unit. For both types of experiments, one pattern (PR, in the field of the right eye in figure 5.18) had constant velocity.
Since the initial motion direction for PL was out, the left unit's response was inhibited during the first half of the pattern presentation—except for a.
PL SUPERIOR PR
OSCILLATING ~ INFERIOR
ALTERNATING 111:1 I>-
OUTWARD (a)
POSITION OF STRIPES
ON OFF
OSCILLATING PL
ALTERNATING PL
VELOCITY
Stimuli for the second type of experiment were similar to those described for the first, except that instead of the direction of a pattern reversing during the on period, the direction was changed when the pattern was off (Figure 5.18). The top two curves of Figure 5.19(b) show the response of the left unit to the alternating pattern, the bottom two are the responses of the right unit to the constant pattern. Of the latter two, the one with the highest response is for the outward movement of the alternating cartridge, that is, for the cycle in which the left unit was inhibited by the cartridge.
For the experiments in which oscillating patterns were used, two intervals were chosen directly adjacent and on either side of the tip of pattern. A general increase in the amount of inhibition with increasing rate of the inhibitory unit was observed. It was then observed that the amount of inhibition appeared to depend not only on the firing rate of the inhibiting unit, but also on the inhibited unit's rate.
Let pD and pD be the velocities of the unit that directly responds to the pattern by oscillating or switching.
In earlier work [58] there was an initial fluctuation in the rate of fire upon sudden presentation of a cartridge. That is, two different cartridges boosting the inhibiting unit should produce the same effect if the rate of fire for both cartridges were the same. The patterns were PR in right field and P LA and P LB in left field.
The first was performed with the stimuli as shown, except that the directions of FLA and PLB were always inward. Under these conditions, the intensities, rates, and position of P LA and P LB were adjusted so that the left unit response was approximately the same for both. Adjustments were made based on the oscilloscope and speaker output, and by monitoring the SCAN screen (pre-analysis).
SUPERIOR PR
INFERIOR
The corresponding four histograms for the inhibited unit were then calculated and the amount of inhibition for PLA and PLB measured as described above. The ratio between the difference in the amount of inhibition and the rate of fire of the inhibited unit varied between 0. The variation in the amount of inhibition suggested that the effect on the inhibited unit might depend on the specific movement patterns in its field.
That is, for model A, it is assumed that if only Pl is present, the rates at point "a" are Pru and 0 for the upper and lower branches. Assume that Pl is presented only under the condition of two different firing rates for the left unit: p L > O; and pL = 0. Trials were approximately equally distributed between the inhibited unit being left or right Ila-in.
The combinations of patterns used to stimulate the inhibited unit (Pl and PZ) are schematically illustrated below.
PL SUPERIOR
IN F ERIOR
P L OFF
RIGHT UNIT RESP~NSES T~ RLTERNRTING PL
OU TH ARD
The experiments considered here are limited to those in which two internal motion detection units having visual fields in the same eye were recorded simultaneously. Explicitly excluded are, for example, combinations such as Ila-in and Ila-out recorded in the same lobe and IIa-in1 recorded in opposite lobes. In no case were spike trains synchronized when both electrodes were recording from class Ila units in the same lobe.
That is, whenever a pair of internal motion detectors recorded in the contralateral lobe, they were synchronized. In the second case, an Ila-in - Ila-out pair was recorded in the same lobe. The correlogram between an Ila-down and Ila-in recorded in the same lobe was flat, indicating little or no interaction.
In one case, no correlation was observed; in the other, a weak inhibitory valley was noted, similar to that for opposite Ila-in's.
CHAPTER VI
For these two pairs, peaks in the cross-correlogram demonstrated the existence of interaction between members of the pairs. Furthermore, none of the results contradict the claim, such as finding an unsynchronized pair of Ila's in the same lobe. In one of the two cases in which opposite Ilbs were recorded, mutual inhibition was observed.
In addition, since the correlogram between inward and outward motion detectors with visual fields in the same eye showed an inhibitory relationship, the effect of the outward unit must also be considered. It is interesting to interpret the existence of both inward and outward motion detectors in light of the previously mentioned optomotor work by Goetz [28] on Drosophila and Musca. That is, inward motion excites one side of the motor system via the branch descending to the subesophageal ganglia, and inhibits the opposite side via the inhibitory effect on the opposite inward motion detector.
That is, four out of the six pairs possible with inward or outward motion detectors with visual fields in the same or opposite eyes were recorded simultaneously (the missing combinations are pairs of Ila-out's that have fields in the same or opposite eyes).
