Note: Descriptions are shown in the official language in which they were submitted.
~117643
APPAR~TUS FO~ MEASURING THE
. _
AXI~L LENGTH OF AN ~YE
_
This invention relates to an apparatus for effec-
ting rapid and accurate ultrasonic measurements of the
axial length of the eye.
A compact and simple ultrasonic instrument (echo-
oculometer) for measuring the axial length and anteriorchamber depth of the eye has been described in the litera-
ture by Mortimer et al. in the Proceedings of the 11th Intl.
Conf. on Medical and Biol. Engng. 1976 pp. 508-509 and by
Mortimer et al. in the Canadian J. Ophthal. Volume 12, 1977
pp. 318-320.
Advantages of this instrument over conventional
A-Scan devices employing cathode ray tubes for display are
that a display of the A-Scan is not essential, the results
are immediately available on a counter and are expressed in
convenient numerical units.
The above echo-oculometer utilizes techniques si-
milar to those used in the echo-encep~alo~raph invented by
Hudson et al. and described in U.S. Patent No. 3,872,858
issued ~arch 25, 1~75 and its corresponding Canadian Patent
No. 973,632 issued August 26, 1975. The echo-oculometer
employs a transducer which emits a short pulse of ultrasound
~117643
aimed along the ocular axis~ The echoes returning from the
various surfaces within the eye are received by ~he same
transducer and the time required for the sound pulse to re-
turn is converted to a length measurement. Two range gates
consisting of electronic logic circuits allow echoes from
particular interfaces to be sel~cted and the corresponding
time interval to be measured~ The retinal echo is selected
for the axial length and the anterior lens echo is selected
for the anterior chamber depth measurement. The statisti-
cal accuracy of the determinations may be increased by averaging several readings.
Three important features of the above echo-oculo-
meter device are that it employs a slow gain sweep, a crys-
tal controlled time base of a particular frequency, and a
special delay circuit which determines the time at which
the counter starts to count.
In contrast to the echo-encephalograph which em-
ploys a second fast gain sweep (functioning analogously to
the time varied gain or TGC of the conventional A and B
scan equipment), the oculometer has only a slow gain sweep
since compensation for tissue attenuation is not needed.
For the slow gain sweep, the gain does not vary significan-
tly during the time a given pulse is transmitted and its
echoes are recei~ed. Rather the receiver gain increases
from a transmitted pulse to the next until the range-gated
signal exceeds a predetermined threshold, stopping the
counter (displaying the accumulated count), or until the
maximum recei~er gain level is reached whereupon the mea-
~li7643
surement cycle ls aut~matically repeated.
The gain sweep has two functions. It i5 supposedto compensate for differences in the amplitudes of the re-
ceived echoes (in different eyes) and, more important, it
is supposed to insure that the strongest echo detected in
the range gate inter~al selected will be the first to ex-
ceed the detection threshold thereby stopping the counter.
Key assumptions made in the operation of the abo-
Ye echo-oculometer device are that (1I when the beam is di-
rected along the axis of the eye, the first echo in theselected range gate will be the largest echo, and (2) if
the beam is directed off axis, the echoes received will be
too weak (due to the inclination of the reflecting interface
and the transducer directivity) to exceed the range gate de-
tection threshold.
While these assumptions are generally true for theanterior lens echoes (,anterior chamber depth), research and
clinical experience conclusively show that this is not al-
ways true for the much more important retinal echoes (axial
length). In a significant number of cases, other interfaces
and structures behind the retina give rise to the largest
echoes. This can result in errors in the axial length de-
terminations of as much as 3mm which corresponds to an error
in the lens power determinations of about 8 or 9 diopters.
An error of this magnitude is altogether unacceptable.
Furthermore, lt must be emphasized that although this pro-
blem is much more common for the case of off-axis beam in-
cidence, it will still sometimes occur when the beam is pro-
1~7~3
perly aligned.
Finally, if readings are taken with the beam im-
properly aligned, the accuracy of the axial length determi-
nations will be decreased either due to the problem just
described or in the case where the retinal echo does stop
the counter, the fact~ that a chord shorter than the axial
diameter is being measured.
One method of counteracting these problems would
bé to increase the beam directivity but this is subject to
both theoretical and practical limitations. Another ap-
proach might be to lower the upper limit of the swept gain.
However, there are ob~ious constraints since the instrument
must accommodate a considerable range of ultrasonic and
geometrical characteristic~ for different eyes~
Another method would be to require the presence
of the anterior and posterior lens echoes (either one but
preferably both) of a magnitude equal or greater than some
specified fraction of the retinal echo threshold (typically
~ or more) as a necessary condition for a valid reading.
In fact, the standard A-scan technique (used in determining
the axial length) consists of insuring that both the ante-
rior and posterior lens echoes are simultaneously present
together with the retinal echo, and then maximizing the two
lens echoes while maintaining a good clean and large retinal
echo. The technique is illustrated in detail by Leary in
Ultrasonics April 1967, pp. 84-87, Under normal conditions,
the symmetry of the eye is such that following the above
procedure will insure good axial alignment.
~lq643
Implementing the abo~e lens echo conditions ln
the form of electronic circuits to assure that readings are
obtained only under conditions of good alignment is straight
forward and is obvious to those skilled in the art. Howe-
ver~ while this would greatly reduce the likelihood of in-
correct ~riggerlng by structures behind the retina, it does
not entirely eliminate the problem. Also the difficulty of
achieving exact alignment without reference to an A-Scan
can make the actual obtaining of readings very problematic.
lQ Certainly adding the lens echo condition reduces the speed
with which valid readings can be obtained. This i5 an im-
portant consideration when dealing with older or uncoopera-
tive patients.
It is therefore the object of the present inven-
tion to provide an apparatus for effecting rapid and accu-
rate ultrasonic measurements of the axial length of the eye.
The apparatus, in accordance with the invention,
comprises a transducer adapted to transmit repetitive ultra-
sonic pulses along the ocular axis of the eye of a patient
and receive echo pulses reflected from the retina of the
eye, a fixed gain amplifier connected to the transducer for
amplifying such reflected echo pulses, an automatic gain
controlled amplifier also connected to the transducer for
amplifying the reflected echo pulses, control means coupled
to the automatic gain controlled amplifier for gradually
increasing the gain of the amplifier during a measurement
cycle, first and second gate circuits controlled by the
output of the fixed and automatic gain controlled amplifier,
11176~3
respectively, and adapted to pass logic signals triggered
by retinal echo pulses exceeding first and second predeter-
mined thresholds, a digital counter connected to the second
gate circuit and adapted to display the axial length of the
eye as a function of the distance travelled by the retinal
echo pulses, a gate dela~ initiated by a slow clock, a gate
width generator connected to the gate delay for generating
a time slot during which echo pulses originating from the
posterior wall of the eye can be received, a latching cir-
cuit responsive to the gate delay for enabling the firstgate circuit to pass logic signals triggered by echo pulses
exceeding the first threshold, a retinal echo triggered gate
width generator interconnecting the first and second gate
circuits and responsive to the first gate circuit for ena-
bling the second gate circuit to pass logic signals trigge-
red by retinal echo pulses exceeding the second threshold in
the time slot generated by the gate width generator, the
output of the retinal echo triggered gate width generator
being also connected to the latching circuit for blocking
the first gate circuit immediately after receipt of the
first logic signal triggered by a retinal echo pulse, there-
b~ preventing mistriggering of the retinal echo triggered
gate width generator by echoes originating from structures
behind the retina~
The above control means includes a slow ramp gene-
rator adapted to generate a ramp voltage which is applied to
the automatic gain controlled ampllfier in such a manner
that the gain of the amplifier varies from a minimum at the
76~3
start of the measurement cycle to a preset maximum after a
number of cycles~
A pulser is connected to the transducer for ap-
plying sharp hlgh voltage splkes to such transducer to shock
excite lt so as to direct an ultrasonic pulse into ~ne eye
being examined. The pulser is triggered by the slow clock~
A fast clock, operating at a frequency in MHz which is l/2
times the average velocity of ultxasound in the eye expres-
sed in units of 0.1 mm per microsecond, provides the coun-
ter time base and is used to synchronize the slow clock.
A pulser delay is located between the slow clockand the pulser to insure that the counter starts counting
at the correct time~
The invention will now be disclosed, by way of
example, with reference to the accompanying drawings in
which:
Figure l is a block diagram of an exemplary embo-
diment of the invention;
Figure 2 is a series of waveforms produced at va-
rious points in the block diagram of Figure l; and
Figure 3 is a circuit diagram of the echo-oculome-
ter constructed according to the present invention.
Referring to Figures l and 2, there is shown a
fast clock 10 which generates a continuous signal at a fre-
quency in MHz which is 1/2 times the average velocity of ul-
trasound in the eye ~1553 m/sec~ in units of 0~1 mm per mi-
crosecond, that is 7.765 MHzt,and has a sinusoidal waveform
as illustrated at A tn Figure 2. The clock lO feeds a si-
1~76~3
gnal to a counter-display 12 and to a slow clock 14. The
slow clock 14 generates a square ~ave at a frequency of
about 60 Hz as shown at B in Figure 2 of the drawings. The
operation of the slow clock 14 is synchronized to the fast
clock 10. The output slgnal of the slow clock is applied to
a pulser delay 1~ which generates a repetitive pulse signal
such as shown at C in Figure 2~ The pulser delay 16 trig-
gers a pulser 18 on the positive going edge of the waveform
C and at the same time resets the counter 12. The counter
begins counting on the negative going edge of the waveform
C to compensate for the propagation time of the ultrasonic
pulse from the transducer to the eye and from the eye to the
transducer. Pulser 18 generates a sharp high voltage spike,
such as shown at D in Figure 2, which is used to shock exci-
te a transducer 2Q to direct an ultrasonic pulse through theeye being examined. This pulse travels through the eye and
is reflected by the various surfaces and returned to the
transducer. These echo pulses are picked up by the transdu-
cer and converted back to electrical signals which are di-
rected to a preamplifier 22 providing an output such asshown at E in Figure 2. The output of the preamplifier 22
is applied to a fixed gain amplifier 24 providing an ampli-
fied output such as shown at F in Figure 2 and to an automa-
tic gain control amplifier 26 providing an amplified output
such as shown at G, H or I in Figure 2 depending on the gain
of the amplifier as controlled by the voltage V applied to
its automatic gain control terminal, The output of the fi-
xed galn amplifier 24 is fed to a first gate circuit 28
1:~L1764~3
through a comparator 30 which sets a signal threshold level
for the output of amplifier 24 as illustrated by a dashed
line through waveform F in Figure 2. Similarly, the output
of the automatic gain control amplifier 26 is fed to a se-
cond gate clrcuit 32 t~rough a comparator 34 which sets asignal threshold level for the output of amplifier 26 as
illustrated by a dashed line through waveform I in Figure 2.
The pr~sent apparatus is capable of measuring not
only the axial length (AL) of an eye by detecting the reti-
nal echo pulses but also the anterior chamber (AC) depth bydetectIng the anterior lens echo pulses. As mentioned pre-
viously, the measurement of the anterior lens echoes is not
the object of the present invention, therefore the portion
of the circuitry which is concerned with the measurement of
the retinal lens echo will be primarily disclosed. In or-
der to permit gate circuits 28 and 32 to pass logic signals
triggered by the echo pulses which are reflected from the
desired surfaces of the eye, there is provided a gate delay
36 which is triggered by the positive leading edge of the
signal B appearing at the output of the slow clock 14. Gate
delay 36 provides an output waveform such as illustrated at
L in Figure 2, and triggers the gate width generator 38 on
the positive edge of the waveform L. If the anterior cham-
ber depth was to be measured, the delay would of course be
much shorter so as to allow gating of the echo pulses origi-
nating from the anterior lens surface of the eye. A func-
tion switch 37 is- provided for selecting which one of the
measurements is to be performed by the apparatus. The gate
~'7ti~3
width generator 38 generates a signal M as shown in Figure
2. Signal M is fed to gate 32 and controls the time slo.
during which gate 32 is opened. Gate width generator 38 is
also responsive to function switch 37 for selection of the
desired measurement to be performed. The output L of gate
delay 36 is also applied to a spike generator 40 which gene-
rates a signal shown at N in Figure 2. Signal N is fed to
a latching circuit 42 which generates a signal P, as shown
in Figure 2, for controlling the opening of gate circuit 28.
Latching circuit 42 is reset by the output C of the pulser
delay at the beginning of each cycle.
A retinal echo triggered gate width generator 44
is connected between gates 28 and 32 and is triggered by
output signal Q of gate circuit 28 when a retinal echo si-
gnal exceeding the threshold of comparator 30 is present.The retinal echo triggered gate width generator 44 provides
an output R, as shown in ~igure 2, which is applied to the
gate circuit 32 to permit the gate to pass logic signals
triggered by echo signals originating from automatic gain
control amplifier 26 exceeding the threshold of comparator
34. The output R is also fed to the latching circuit 42 to
cause the latching circuit to immediately disable gate
circuit 28 after receipt of the first logic signal triggered
by an echo signal originating from the fixed gain amplifier
24. Thus, gate circuit 28 is latched out immediately after
the retinal echo pulses are detected to prevent retriggering
of 44 thereby pre~enting gate 32 from passing logic signals
triggered by echo pulses originating from structures behind
1~76~3
11
the retina~ The retinal echo triggered gate width genera-
tor 44 is dlsabled by the function switch 37 during anterior
chamber measurement because it is not required.
The output T of gate 32 is fed to a display dura-
S tion circuit 46 whic~ provides an output U to stop the coun-
ter and display,for a few seconds, the distance travelled by
the retinal echo pulse as an indication of the axial length
of the eye.
The gain of the fixed gain amplifier 24 is set by
an amount approximately 10 to 14 dB greater (3 to 5 times
greater) than the maximum gain of the automatic gain control
amplifier 26, This insures that the retinal echo signal
will be of su~ficient amplitude to exceed the threshold of
the comparator 30 and that the retinal echo triggered gate
width generator 44 will not be mistriggered by an echo pulse
of greater amplitude originating from structures behind the
retina.
The gain of the automatic gain control amplifier
26 is varied by a slow ramp generator 48 which generates a
voltage of increasing negative amplitude V starting from a
minimum value at the beginning of the measurement up to a
maxim~m value set by a comparator 50. The output U of the
display duration circuit 46 is applied to the slow clock to
enable the same, and to the slow ramp generator 48 to reset
the slow ramp generator voltage V to its mlnimum value, when
the automatic gain control amplifier has sufficient gain to
pass the signals exceeding the threshold set by comparator
34~
ill76~3
12
The invention will now be disclosed with reference
to the more detailed circuit diagram of Figure 3 which is
intended to give a better understanding of the invention but
not to limit t~e scope thereof. The non-detailed blocks as
well as the circuit diagrams outlined in Figure 3 by broken
lines carry the same references as the corresponding blocks
of Figure l.
The fast clock 10 is a conventional crystal con-
trolled oscillator operating at a frequency of 7.765 MHz as
mentioned previously. The output A of the fast clock is fed
to the counter-display 12 which is a conventional digital
counter capable of displaying a count when energized to do
so. A suitable example of such a counter is RCA No.
ICAN-6733. The counter-display is therefore operated by the
fast clock to indicate directly the axial length of the eye.
For synchronizing of the clocks, the output A of the fast
clock is also applied to the clock input C of a conventional
type D flip~flop 60 which acts as a slow clock. The timing
period of the slow clock is about 60 Hz as mentioned pre-
viously and is determined by a resistor Rl connected between
terminals D and Q of the flip-flop and capacitor Cl connec-
ted between terminal D and ground. Flip-flop 60 may be ena-
bled by clamping terminal D through diode Dl as it will be
disclosed later.
The output of the slow clock is applied to pulser
delay 16 through coupling capacitor C2 and resistor R2. The
pulser delay is comprised of a CMOS NOR gate 62 and an in-
~rerter 64. Gate 62 has a first input connected to the slow
1~76~3
clock and a second input connected to the output of inver-
ter 64~ The output of gate 62 is connected to the input of
the inverter 64 through capacitor C3. A positive potential
V+ is also applied to t~e input of the inverter through re-
sistors R3 and R4~ Gate 62 and inverter 64 form a wellknown monostable circuit. The output of the pulser delay
i5 as shown at C in Figure 2 of the drawings.
The pulser 18, which is energized from a conven-
tional high voltage source 66, is triggered on the rising
edge of output C of the pulser delay and produces a sharp
high voltage spike D which is used to shock excite the
transducer 20 to direct an ul~rasonic pulse into the eye of
the person being examined. This pulse travels through the
eye and is reflected by various surfaces of the eye, as men-
tioned previously, and returns to the transducer. The echopulses are detected by the transducer and converted back to
electric signals which are fed to preamplifier 22. The out-
put of preamplifier 22 is applied to a fixed hi-gain ampli-
fier 24 and to an automatic gain control amplifier 26. The
above mentioned circuit elements 18, 22, 24 and 26 are con-
ventional and need not be disclosed in detail.
The output F of fixed hi~gain amplifier 24 is
full-wave rectified by diodes D2 and D3 and clamped to the
voltage level determined by the voltage d vider resistors
R5, R6 connected across a source V ; The clamped signal
~Gate trig2 is fed to one of the inputs of a two input NOR
gate 68. T~e echo signal detection threshold voltage is
equal to the difference between the NOR gate logic threshold
1~7643
14
and the above clamping voltage. In this particular embodi-
ment, therefore, the resistor network R5 and R6 and the lo-
gic threshold (approximately ~V+ for CMOS logic) constitute
essentially the equivalent of the comparator 30 while the
S NOR gate 68 constitutes the gatlng circuit 28.
Resistors R7 and R8 and capacitor C4 provide lo-
pass flltering for the echo signals which improves the de-
tection performance. While not essential, full-wave recti-
fication of the signal simplifies lo pass flltering of the
signal (in order to improve the signal to noise).
In a manner similar to the preceeding, the output
(G, H, I) of the automatic gain control amplifier is clam-
ped to the voltage level determined by the voltage divider,
resistors R9 and R10, connected across the V+ source. ~he
clamped signal (stop) is fed to one lnput of the NOR gate
circuit 70. R9 and Rl0 and the NOR gate logic threshold
thus form the equivalent of the comparator 34, and the three
input NOR gate 70 corresponds to the gating circui~ 32.
The gain of the amplifier 26 is Yariable and de-
pends on the "sweep" voltage applied to its AGC terminal asit will be disclosed later.
A diode D4 is connected across the resistor R9 to
protect the CMOS against overvoltage.
The output B of the slow clock 14 is also applied
to the gate delay 36. Gate delay 36 comprises a C~OS ~OR
gate 72 and an inverter 74 which are interconnected in th~
same manner as in pulser delay 16 to form a monostable.
Gate 72 has a first input connected to the slow clock 14
1117643
and a second input connected to the output of invertex 74.
The output of gate 72 is connected to the input of the in-
verter 74 through a capacitor C5. A posltive potential V+
is also applied to the input of the inverter through resis-
tors Rll-R14. The time constant of the monostable is con-
trolled by resistors Rll-R14 and capacitor C5. The output
L of the monostable is as shown in Figure 2 of the draw-
ings. Gate delay 36 may also be used for anterior chamber
-- measurement and, in such a case, the time constant of the
lQ R-C circuit may be changed by clamping the connecting point
of resistors R12 and R13 to the voltage source V'+ through a
diode D5. The clamping action is performed by operating a
switch in function switch 37 (Figure 1). As mentioned pre-
viously, the purpose of the gate delay 36 is to delay the
operation of the gate width generator 38 which sets the time
slot during which the desired echo is to be detected.
The output L of the gate delay 36 is applied,
through a capacitor C6, to the gate width generator 38 which
comprises a CMOS NOR gate 76 and an inverter 78. The gate
width generator is a monostable of the same type as the one
of the pulser delay 16 and the gate delay 36. The first in-
put of the gate 76 is connected to output L of the gate de-
lay and its second input is connected to the output of in-
verter 78. The output of gate 76 is connected to the input
of inverter 78 through a capacitor C7. A positive potential
source V~ is also connected to the inpu~ of inverter 78
through resistors R16-Rl9. The time constant of the mono-
stable is set by resistors R16-Rl9 and capacitor C7. The
1117643
16
output M of the monostable is as shown in Figure 2 of the
drawings~
The output M of the gate width monostable 38 de-
termines the length of time during which the echo returned
from the posterior part of the eye will be detected. Such
output M is applied to the gate circuit 32 as it will be
disclosed later. The gate width generator 38 may also be
used for anterior chamber measurement and, in such a case,
the time constant of the R-C circuit will be chanyed by
clamping the connecting point of resistors R17 and R18 to
V'~ through a diode D6. This is performed by function
switch 37.
The output L of the gate delay 36 is also applied
to a latching circuit 42 through a spike generator 40 formed
by resistor R15 and capacitor C6. In the embodiment disclo-
sed, latching circuit 42 is a conventional type D flip-flop
80~ The output N of the spike generator is applied to the
reset terminal of flip-flop 80 to reset the output of the
flip-flop to low at the beginning of the time slot during
which theechoes returned from theposterior part of the eye
are to be detected r as indicated by waveform P in Figure 2
of the drawings.
The output P of the latching circuit is applied to
the CMOS NOR gate 68 to enable the gate. When no echo signal
~amplified by fixed gain amplifier 24) exceeds the threshold
set by comparator 30r the output Q of gate 68 remains low.
CMOS NOR gates 68 and 70 are interconnected by a
retinal echo triggered (R.E.T.l gate width generator 44 com-
~1176~;~
17
prising a CMOS NOR gate 82 and an inver~er 84 which operateas a monostable~ Gate 82 has a first input connected to
ground through resistor R20, a second input connected to the
output Q of gate 68 and a third input connected to the out-
put of in~erter 84~ The output of gate 82 is connected tothe input of inverter 88 through capacitor C8. The lnput of
the inverter 84 is also connected to a positive potential
source Vl through resistor R21. The inverter is protected
against overvoltage by diode D7. The input of the inverter
is normally high as it is connected to V~ and its output
low, so that when the output Q of gate 68 is low (no echo
signal exceeding the threshold of comparator 30), the output
R of gate 82 is high. However, when an echo signal excee-
ding the threshold level is received, the output Q of gate
68 turns high and the outpu~ R of gate 82 turns low. The
output R o~ gate 82 is applied to terminal C (clear) of
flip-flop 80 to turn output Q of the flip-flop high to imme-
diately block gate 68 and so prevent the R.E.T. gate width
generator from being retriggered after it has been triggered
2Q by a retinal echo signal. Thus, output Q is only a narrow
spike such as shown in Figure 2 of the drawings. After ap-
proximately one microsecond as set by resistor R21 and capa-
citor C8, the output of inverter 84 returns to low and the
output of gate 82 to high thereby blocking gate 70. As a
result, any subsequent echos which exceed the threshold of
comparator 34 are prevented from trig~ering gate 3~ If no
retinal echo s~gnal exceeding the threshold level of the
comparator 30 is detected, flip-flop 80 is set ~Q = 1) by the
1117643
18
output C of pulser delay 16 at the beginning of the next
cycle to block gate 68 and so prevent the latching circuit
from being activated by echo signals detected in the gate
delay interval which would disable the R.E~T. gate width
generator prematurely. The R ~ E ~ T ~ gate width generator may
also be disabled by applying a positive voltage ~'+ to the
first input of gate 82. This may be done by a switch of
function switch 37 when making an anterior chamber measure-
ment as the R.E.T. gate width generator is not needed for
such an operation.
When a retinal echo pulse is detected, the output
of the R.E.~. gate width generator is applied to the first
input of the three input NOR gate 70. As long as the reti-
nal echo signal amplified by the automatic gain control
(AGC) amplifier does not exceed the threshold set by compa-
rator 34, the clamped output signal stop applied to the se-
cond input of gate 70 is logical~ ~ig~ and the output T o~
ga~a 7Q is low~
As mentioned previously, the gain of the AGC am-
2Q plifier is controlled by a slow ramp generator 48 and a com-
parator 50. The slow ramp generator i5 a conventional Mil-
ler integrator comprising a resistor-capacitor charging net-
work consisting of resistor R23 connected to the inverting
terminal of an operational amplifier 86 and a capacitor C10
connected between the inverting terminal and the output ter-
minal of the operational amplifier. The output "Sweep" of
the operational amplifier provides a linear time~base volta-
ge varying from a minimum voltage of say 5V to a maximum vol-
1~76~3
19
tage of say 12V under the control of comparator sa~ Thecomparator 50 comprises an operational amplifier 88 having
its inverting terminal connected to source V+ through resis-
tors R24 and R25 and its inverting terminal connected to the
junction of resistors R26 and R27 which are connected in se-
ries with a diode D10 between the "sweep" output of the ope-
rational amplifier 86 and ground. A resistor R28 is connec-
ted between the non-inverting terminal and the output termi-
nal of the operational amplifier 88 for controlling the gain
lQ thereof in known manner. The output of operational ampli-
fier 88 is connected to the non-inverting terminal of the
operational amplifier 86 through a coupling resistor R29.
The "sweep" output of the ramp generator is shown at V in
Figure 2 of the drawings but not on the same time scale as
the other waveforms. The time lapse between tl and t2 is
about 0.7 sec and between tS and t6 about 2 sec. The ampli-
tudes of the echo signals at times t3, t4 and t5 are shown
at G, H and I in Figure 2 of the drawings. At the beginning
of the measurement cycle, the gain of the AGC amplifier 26
is low but it gradually increases until at time t5 the reti-
nal echo exceeds the threshold set by comparator 34. At
such time, the "Stop" input of gate 70 turns low and, if the
other two inputs of gate 70 are also low (in the time slots
generated by the gate width generator 38 and the R~E~T. gate
width generator 44~, the output T of gate 70 turns high.
The output T of gate 70 is applied to a display
duration monostable circuit comprising a CMOS NOR gate 90
and inverter 92. ~he first input of gate 90 i5 connected
11~76~3
to the output of gate 70 and its second input is connected
to the output of the in~erter 92. The output of the gate
90 is connected to the input of inverter 92 through a capa-
citor Cll. A positi~e potential source v~ is also connected
to the input of inverter 92 through resistor R30. A protec-
tive diode Dll is connected across resistor R30 to protect
the inverter agai~st overvoltage. The output of the mono-
stable is shown at U in Figure 2 of the drawings. The out-
put signal U is applied to the "Display enable" terminal of
the coun~er and display 12 to stop the counter and show the
axial length of the eye for a period of time determined by
the time constant of resistor R30 and capacitor Cll~ The
output U of the inverter 92 is applied to an inverter 94
which produces the output "Display enable" which is applied
to the slow clock 14 to enable the slow clock. The output
"Display enable" is also applied to the inverting terminal
of operational amplifier 86 through diode D12 to reset the
gain sweep ramp voltage generator 48.
It will be seen from the above description that
the counter 12 will be stopped and the display turned on on-
ly if (1) the output of the automatic gain amplifier 26 ex-
ceeds the threshold set by the comparator 34 and if (2) this
occurs during the short time interval that the gate circuit
32 ~or CMOS NOR gate 70) is enabled by the R.E.~. gate width
generator 44 as indicated by waveform R in Figure 2 of the
drawings. If the beam is off-axis, the retinal echo will
not have enough amplitude to stop the counter (which is then
automatically reset on the next pulse transmitted). As a
11~76~3
result, the echo-oculometer in accordance with the present
invention behaves as if the beam directivity was much grea-
ter than it actually is. More importantly, mistriggering
by echoes from structures behind the retina is eliminated by
the latching circuit which prevents the R~E.T. gate width
generator from ~eing retriggered by echoes from structures
behind the retina after it has been triggered by the retinal
echo pulses~ It is also important to note that the signal E
appearing at the output of fixed gain amplifier 24 is ampli-
lQ fied by an amount approximately 10 to 14 dB greater than themaximum gain of the AGC amplifier 26 to make sure that the
first echo signal exceeding the threshold of the comparator
30 is positively a retinal echo signal and not an echo si-
gnal originating from a structure behind the retina as would
happen if the retina echo was of lower amplitude than the
threshold set by the comparator 30.
Although the invention has been disclosed with re-
ference to a workable embodiment shown in Figure 3 of the
drawings, it is to be understood that other detailed circuit
diagrams could be used for the blocks of Figure 1 and that
the invention is not limited to such detailed circuit dia-
grams.
