JPH06114077A - Voice controller for microscope for operation - Google Patents

Abstract

PURPOSE:To provide a voice controller for an operation microscope which allows the heightening of a recognition rate regardless of changes in the feature of voice of a user and is excellent in operability. CONSTITUTION:A microscope for operation with drive means 17 and 18 is provided with a voice controller 11 having a voice recognition means 13 and the drive means 17 and 18 are operated by voice. The voice controller 11 has a means 14 which computes a reference pattern data of operation directives of the voice recognition means 13 registered beforehand and a pattern data of an input voice to renew the contents of the reference pattern data based on the results of the computation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an operation input device for each drive unit of a surgical microscope used in the medical field.

[0002]

2. Description of the Related Art In recent years, with the development of surgical techniques and surgical tools, so-called microsurgery, which is a surgical operation on a minute affected area, has been frequently performed. As a microsurgery, a surgical microscope for magnifying and observing a surgical site is used, for example, as seen in neurosurgery and orthopedic / plastic surgery. In microsurgery,
In addition to surgical forceps such as scissors, tweezers, and scalpel, the surgeon may use surgical instruments such as bipolar and electric scalpel that are operated by a foot switch. In such a case,
Both hands and feet are often blocked.

Further, since the operation inputs such as the magnification change, focus adjustment and the movement of the body of the surgical microscope are usually performed by the foot switches, the number of foot switches is increased,
There has been a problem that it becomes difficult to operate various foot switches quickly and accurately during surgery. Therefore, there is an increasing need to perform each operation input of the surgical microscope, which is conventionally performed by the foot switch, by using a voice controller capable of automatically controlling each drive unit by voice recognition based on a voice uttered by an operator or the like. There is.

In the voice recognition device built into the voice controller, a specific story in which the user registers in advance the standard pattern data of the operation instruction word of the user in the RAM of the voice recognition device before performing the voice recognition. There is a known speaker system and an unspecified speaker system in which standard pattern data extracted in advance from a sample of operation instruction words for dozens of people is written in the RAM of the voice recognition device and a user need not be registered. There is. However, since the voice controller used in the above-mentioned surgical environment is required to have high reliability and safety, and the number of operation instruction words is as small as about a dozen words, it is easy to obtain a relatively high recognition rate. Speaker-type voice recognition devices have become mainstream.

Further, the recognition means of the voice recognition device generally compares the standard pattern data of a plurality of operation instruction words registered in advance with the pattern data of the operation instruction voice inputted from the microphone, and makes the pattern data similar to each other. It is known that a degree point is calculated using a similarity degree method such as a simple similarity degree method, and operation instruction data corresponding to an operation instruction word having pattern data with the highest degree of similarity (similarity point) is output. There is. However, the utterance sound of the operation instruction word uttered by the user is not always constant, and even if the utterance sound is the same, the standard pattern data of the utterance sound is registered due to unstable factors such as physical condition and mood of the day. Subtle changes in the voice characteristics of the uttered sound often occur in comparison with time.
Therefore, the degree of similarity with the input standard pattern data becomes low, which may lead to a reduction in the recognition rate.

In order to avoid this, it is necessary to re-register the standard pattern data at an appropriate time, and it is preferable to re-register the standard pattern data immediately before using the voice controller to obtain a high recognition rate. However, re-registration is extremely troublesome, and it is troublesome because registration can be performed only in the vicinity of a surgical microscope having a built-in voice controller. In addition to this, depending on the progress of the operation even during the operation, for example, when an emergency treatment such as sudden bleeding has to be performed, due to the mental upsurge of the operator, the voice characteristics may change before the operation. There is a risk that the recognition rate will decrease, as it may change compared to.

Further, in the case of an emergency operation or the like, there is no time to re-register the standard pattern data, and an operation using a voice controller with a low recognition rate is inevitably performed, which only prolongs the operation. Moreover, there is a risk that the safety of the surgery will be impaired and the surgeon will be severely mentally burdened.

As an example of applying voice recognition to a surgical microscope, according to Japanese Patent Laid-Open No. 62-166312, voice recognition is used for mode switching of a foot switch of a surgical microscope, and the number of contacts of the foot switch is reduced. In addition, an example in which the voice recognition device itself is supplemented is disclosed.

[0009]

However, in the above-mentioned conventional example, since the voice recognition is used to switch the mode of the foot switch of the surgical microscope, there is a problem that the foot switch must be operated after all. . In addition, since the recognition performance of the voice recognition device itself is not improved, when the voice characteristics of the user change, the recognition rate decreases, not only prolonging the operation time but There was a fear of a great mental burden.

In view of the above situation, an object of the present invention is to provide a voice controller for a surgical microscope which can improve the recognition rate even if the voice characteristics of the user change, and is excellent in operability. To do.

[0011]

A voice controller for a surgical microscope of the present invention is provided with a voice controller having a voice recognition means in a surgical microscope having one or more driving means, and the voice is driven by the voice controller. In the surgical microscope in which the means is operated, the voice controller calculates standard pattern data and pattern data of the input voice of each operation instruction word of the voice recognition means registered in advance, and based on the calculation result. Means for updating the contents of the standard pattern data.

[0012]

According to the present invention, even if the operator's voice is slightly changed after a certain time has elapsed since the operator first registered the standard pattern data of the operation instruction word, the standard pattern data Will be updated. As a result, it is possible to obtain a high recognition rate for a long period of time without re-registering new standard pattern data before surgery, and thus it is possible to greatly reduce the time and effort required for surgery, and in addition, perform emergency surgery. Even in such cases, it is possible to effectively prevent the recognition rate from decreasing.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of a voice controller for a surgical microscope of the present invention will be described below with reference to FIGS. 1 is an overall view of the surgical microscope, and FIG. 2 is a block diagram of the voice controller. In FIG. 1, a microscope body 2 of a surgical microscope 1 is attached to a free end of an arm portion 4 pivotally attached to an upper end portion of a pedestal 3. The arm part 4 is
The first rotatably held around the axis A with respect to the gantry 3.
An arm 5 and a second arm that is rotatable with respect to the first arm 5 about an axis B and is rotatable with respect to a member 6 having the axis B about an axis C (perpendicular to the paper surface). 7 and 7. Thereby, the mirror body 2 can be moved to a desired position in the three-dimensional space.

Further, the mirror body 2 incorporates an electric focus adjusting mechanism and an electric zoom mechanism, which are not shown, and respective control units of the electric focus adjusting mechanism and the electric zoom mechanism are incorporated in the mount 3. . A microphone 9 is fixed to the tip of a microphone arm 8 attached to the mirror body 2. The microphone 9 is connected to a voice controller 11 fixed to the pedestal 3 via a microphone wire 10, and the acoustic signal collected by the microphone 9 is
It is adapted to be transmitted to the voice controller 11 via the microphone wire 10.

In the voice controller 11 shown in FIG. 2, the other end of the microphone wire 10 whose one end is connected to the microphone 9 is connected to the voice recognition device 1 via the amplifier 12.
3 is connected to the voice input unit, and further the voice recognition device 1
Reference numeral 3 denotes an 8-channel band-pass filter (not shown) (not shown) for frequency-analyzing the input voice, and an A / D converter (not shown) in order to create pattern data of an acoustic signal input by the microphone 9. Is equipped with. Further, in the voice recognition device 13, a RAM 50 and a RAM 51 are built in as a voice recognition memory, standard pattern data for each operation instruction word is registered in the RAM 50 as an operation instruction voice dictionary, and the RAM 51 has a voice. It serves as a work area for recognition, and stores pattern data and the like of input acoustic signals.

The controller unit 14 of the voice recognition device 13 outputs a control signal of the drive control unit on the basis of the recognition result of the voice recognition device 13, and is parallel to the voice recognition device 13 by a parallel I / F. It is connected and exchanges data such as various commands and recognition results. The control signal output from the signal output unit of the controller unit 14 is transmitted to each drive control unit built in the gantry 3 via the decoder 15 and the interface cable 16, that is, the electric focus adjustment mechanism control unit 17 and the electric zoom mechanism control unit 18. They are connected to each other and driven and controlled according to the signal input from the controller unit 14.

The voice controller for a surgical microscope of the present invention is constructed as described above, and its operation will be described below. The operator registers a plurality of operation instruction words in the RAM 50 in advance. At this time, each operation instruction word corresponds to each operation signal in advance, and a number is assigned to each operation instruction word.

Next, when the operator utters a voice for operating instructions, for example, "zoom up", while the voice controller 11 is waiting for input of the voice for recognition, this is input to the microphone 9 as an acoustic signal, It is transmitted to the voice controller 11 via the microphone wire 10. The input signal is amplified by the amplifier 12 and then input to the voice recognition device 13, frequency-analyzed by the 8-channel BPF (not shown), and then converted into digital data by the A / D converter (not shown). It

Next, in the voice recognition device 13, the RA
Each standard pattern data in the operation instruction voice dictionary registered in M50 and the RAM input from the microphone 9
The pattern data of the audio signal stored in 51 are sequentially compared using the simple similarity method, and each similarity point is calculated. Further, the voice recognition device 13 outputs to the controller unit 14 the numbers of the operation instruction words corresponding to the pattern data of the second highest rank among the calculated similarity points and the similarity points thereof.

Further, at this time, the operation instruction word having the highest similarity point (in this example, "zoom up")
The number of the operation instruction word corresponding to is also stored in the RAM 51. When the controller unit 14 receives the number of the operation instruction word and the data of the similarity point thereof, if the difference between the similarity points of the two exceeds the preset first value, the similarity point of the higher similarity point is determined. A control signal corresponding to the operation instruction word, that is, an up signal corresponding to the electric zoom mechanism controller 18 is output in this example. The output control signal is output to the decoder 15 and the interface cable 1.
Electric zoom mechanism control unit 1 built in the pedestal 3 via 6
8 and drives an electric zoom mechanism (not shown).

Further, in the controller section 14, the difference between the similarity points corresponding to the above-mentioned top two pattern data input from the voice recognition device 13 is higher than the preset first value. If the value exceeds the value, the voice recognition is performed so that the standard pattern data in the RAM 50 corresponding to the number of the operation instruction word stored in the RAM 51 in the voice recognition device 13 is updated with the pattern data stored in the RAM 51. When the voice recognition device 13 outputs the command to the device 13 and receives the command, the voice recognition device 13 updates the standard pattern data. Thus, the new standard pattern data is held in the RAM 50, and the voice controller 11 is again in the waiting state for inputting the recognition voice.

As described above, even when the operator's voice slightly changes after a certain time has elapsed since the operator's operation instruction word standard pattern data was first registered, the standard pattern data is Will be updated. As a result, it is possible to obtain a high recognition rate for a long period of time without re-registering new standard pattern data before surgery, and thus it is possible to greatly reduce the time and effort required for surgery, and in addition, perform emergency surgery. Even in such cases, it is possible to effectively prevent the recognition rate from decreasing. If the standard pattern data of the operation instruction words is created for each operator, the standard pattern data can be updated for each operator by selecting the standard pattern data of the operator who uses the voice controller. Therefore, even if the operator is changed during the operation, the recognition rate does not decrease.

Next, a second embodiment of the voice controller for a surgical microscope of the present invention will be described with reference to FIG. 3 by using the same reference numerals for the substantially same members as the first embodiment. Note that FIG. 3 is a block diagram of the voice controller. In the voice controller 19, one end is the microphone 9
The other end of the microphone wire 10 connected to is connected to the amplifier 20. The amplifier 20 has outputs of two systems, one of which is connected to the voice recognition device 13 via the BPF 21, and the other of which is connected to the controller unit 14 via the A / D converter 22. There is.

In the second embodiment, when the operator utters a voice for operating instructions, the voice is input to the microphone 9 as an acoustic signal and is input to the amplifier 20 incorporated in the voice controller 19 via the microphone wire 10. After being amplified, it is branched into two outputs. One of the outputs of the two systems is frequency-analyzed by the BPF 21,
It is input to the voice recognition device 13, and voice recognition processing is performed there. The output of the other system is input to the 8-bit A / D converter 22, the power spectrum value is converted into a digital signal, and then the data is output to the controller unit 14.

When the operator speaks, the power spectrum value of the acoustic signal input from the microphone 9 increases. In the controller section 14, a built-in counter (not shown) interrupts every 1/100 second to read the data of the power spectrum value from the A / D converter 22,
It is calculated whether or not it exceeds a preset threshold level.

If the power spectrum value exceeds the threshold level and the control signal is output from the controller unit 14, the controller unit 14 outputs the control signal.
Temporarily suspends the output of the control signal. When the control signal output from the controller unit 14 is temporarily stopped, the pattern data of the acoustic signal corresponding to the power spectrum value exceeding the threshold level is used by the simple similarity method with the standard pattern data stored in the RAM (not shown). Are compared, and the number corresponding to the operation instruction word having the highest similarity point is output as a recognition result from the voice recognition device 13 to the controller unit 14, and the controller unit 14 outputs a new control signal corresponding to the recognition result. I do. At this time, the controller unit 14 stops the output of the control signal that has been temporarily stopped, and restarts the interrupt to the A / D converter 22 at intervals of 1/100 seconds in order to monitor the power spectrum value. Further, when the recognition result is reject (a state in which the standard pattern data corresponding to the input acoustic signal is not stored in the RAM and cannot be recognized), the controller unit 14 controls the control which has been temporarily stopped. Restart signal output.

According to the second embodiment, for example, when a stop command is input to the microphone while the electric focus adjusting mechanism or the electric zoom mechanism (not shown) is driven, it takes about 1/100 seconds after the operator starts speaking. Since the drive of each control unit is temporarily stopped, the above-mentioned drive units do not overrun.
Further, even if the operator makes a voice other than for the operation instruction for exchanging the surgical forceps while each control unit is being driven, the recognition result for the input acoustic signal is rejected.
During that time, only the driving is stopped, the driving thereafter is continued, and it is not necessary to utter the operation instruction voice again. In addition, the interrupt operation performed every 1/100 seconds from the controller unit 14 to the A / D converter 22 to monitor the power spectrum value of the acoustic signal is performed by the controller unit 1
4 may be configured to perform only when a control signal is output to at least one of the control units. In this case, when the controller unit 14 is on standby, it is not necessary to perform an interrupt calculation, so the calculation efficiency is improved.

Next, a third embodiment of the voice controller for a surgical microscope of the present invention will be described with reference to FIG. 4 is a block diagram of the voice controller. A registration block 25 is detachably attached to the voice controller 26 via a connector (not shown).
In the registration block 25, the registration microphone 23 is connected to the voice analysis unit 27 by the microphone wire 24. The voice analysis unit 27 further has a RAM via the control unit 29.
28 is connected to the voice analysis unit 27 and the RAM.
The control unit 28 is configured to be controlled via the control unit 29. The control unit 29 is the voice controller 2
6 is also connected to the controller unit 14 so that data can be exchanged. Further, the registration block 25 is provided with a data transmission switch (not shown).

In the third embodiment, when the operator wants to perform the registration work, the registration block 25 is removed from the voice controller 26 and a registration operation instruction voice is uttered to the registration microphone 23. The acoustic signal is collected by the registration microphone 23 and input to the voice analysis unit 27 via the microphone wire 24. In the voice analysis unit 27, after the input acoustic signal is amplified and frequency analysis is performed, A
It is converted into digital data by the / D conversion. The acoustic signal that has been subjected to the above-described processing in the voice analysis unit 27 is converted into R through the control unit 29 as standard pattern data of operation instruction voice.
It is stored in the AM 28.

When the operator finishes speaking the predetermined registration operation instruction voice, the registration block 25 is moved to the voice controller 26.
When the data transmission switch (not shown) is pressed, the standard pattern data of the operation instruction voice stored in the RAM 28 is sent to the controller section 14 of the voice controller 26 via the control section 29. Further, the controller unit 14 stores the received standard pattern data in the RAM 50 built in the voice recognition device 13.

According to the third embodiment, when the standard pattern data is registered, only the registration block 25 of the voice controller 26 can be removed and the voice for registration can be uttered. It is not necessary to perform the registration work near the fixed microscope base 3 for surgery, and the registration work can be performed in a free place of the operator. Furthermore, if the operator suddenly needs to be replaced, the registration work can be performed without interrupting the use of the voice controller 26, and there is no fear that the operation will take a long time. In addition, the registration block 25 is set to the voice controller 2
There is no problem even if the configuration is such that the registration work can be carried out while it is attached to 6. In addition, although the registration block 25 is detachable from the voice controller 26 in this embodiment, the registration block 2
5 may be configured separately, and only the recording medium such as a memory card in which the standard pattern data is stored may be replaced. In this case, the same effect as above can be obtained.

Next, a fourth embodiment of the voice controller for a surgical microscope of the present invention will be described with reference to FIGS. 5 is an overall view of the surgical microscope according to the fourth embodiment. A voice controller 33 is attached to the surgical microscope mount 3, and a foot switch 30 is connected to the voice controller 33 via a foot switch cable 31. FIG. 6 is a block diagram of the voice controller 33. The voice recognition device 13 built in the voice controller 33 is connected to the controller unit 34 via a parallel I / F, and the controller unit 34 is further connected to the decoder 15. In addition, the foot switch 30 is the foot switch cable 3
1 via the controller unit 3 of the voice controller 33
4 and the timer 32 is connected to the controller unit 34.

In the fourth embodiment, when the seesaw switch for zoom adjustment (not shown) of the foot switch 30 is stepped in the zoom-up direction as shown in FIG.
Is to drive the zoom mechanism (not shown) in the up direction while stepping on the foot switch as shown in Fig. 7B.
A control signal is generated for the electric zoom mechanism controller 18. At this time, the controller unit 34 is not affected by the count of the timer 32.

The operator makes a "fine movement" toward the microphone 9.
Then, the controller unit 34 causes the voice recognition device 13 to
Upon receiving the recognition result from, it switches to the fine movement mode. When the controller unit 34 enters the fine movement mode, when the operator steps on the zoom adjusting seesaw switch of the foot switch 30 in the zoom-up direction as shown in FIG. 7C, the controller unit 34 starts monitoring the count of the timer 32. Then, when the seesaw switch (not shown) is stepped on and detected, the motor is driven to drive the zoom-up mechanism (not shown) in the up direction for 1 second counted by the timer 32 from the start as shown in FIG. A control signal is output to the zoom-up mechanism control unit 18. This operation is the same for zoom down and focus up / down. In addition, the fine movement mode is canceled when the foot switch is not input for 5 seconds or more,
You will be able to enter the usual foot switch.

According to the fourth embodiment, when the operator wants to control the zoom and the focus in a large range, he can depress the pedal of the foot switch 30 to perform the same operation as a normal foot switch operation. Further, when the operator wants to control fine zoom and focus, by switching the controller unit 34 of the voice controller 33 to the fine movement mode by voice input, by pressing the pedal of the foot switch 30, each unit can be driven for only 1 second. Therefore, fine adjustment is easy.

In the above embodiment, the fine adjustment time by the timer 32 is set to 1 second, but it is not limited to this and any time can be set. In particular, it is possible to shorten the fine focus adjustment time when the zoom magnification is high, and to extend the fine focus adjustment time when the zoom magnification is low. Further, since the fine movement mode is automatically released by not operating the foot switch for 5 seconds or more, it is possible to save the trouble of releasing the fine movement mode.

[0037]

As described above, according to the present invention,
By providing a means for calculating the standard pattern data and the pattern data of the input voice for each operation instruction word registered in advance in the voice controller and updating the contents of the standard pattern data based on the calculation result, It is possible to realize a voice controller for a surgical microscope with improved operability and improved recognition rate even if voice characteristics change.

[Brief description of drawings]

FIG. 1 is an overall view of a surgical microscope using a first embodiment of a voice controller for a surgical microscope of the present invention.

FIG. 2 is a block diagram of a first embodiment of a voice controller for a surgical microscope of the present invention.

FIG. 3 is a block diagram of a second embodiment of the voice controller for a surgical microscope of the present invention.

FIG. 4 is a block diagram of a voice controller for a surgical microscope according to a third embodiment of the present invention.

FIG. 5 is an overall view of a surgical microscope using a fourth embodiment of the voice controller for a surgical microscope of the present invention.

FIG. 6 is a block diagram of a voice controller for a surgical microscope according to a fourth embodiment of the present invention.

FIG. 7 is a diagram showing an operation sequence of a foot switch in a fourth embodiment of the voice controller for a surgical microscope of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Operation microscope 2 Mirror body 3 Frame 4 Arm part 5 1st arm 7 2nd arm 8 Microphone arm 9 Microphone 10 Microphone wire 11 Voice controller 12 Amplifier 13 Voice recognition device 14 Controller section 15 Decoder 17 Electric focus adjustment mechanism control section 18 Electric zoom mechanism control unit 19 Voice controller 20 Amplifier 22 A / D converter 23 Microphone for registration 24 Microphone wire 25 Registration block 26 Voice controller 27 Voice analysis unit 28 RAM 30 Foot switch 32 Timer 33 Voice controller 34 Controller unit

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] November 5, 1992

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0004

[Correction method] Change

[Correction content]

In the voice recognition device built into the voice controller, a specific story in which the user registers in advance the standard pattern data of the operation instruction word of the user in the RAM of the voice recognition device before performing the voice recognition. Method and an unspecified speaker method in which standard pattern data extracted in advance from a sample of operation instruction words for dozens of people is written in the ROM of the voice recognition device and the user does not need to register the standard pattern data. It has been known. However, since the voice controller used in the above-mentioned surgical environment is required to have high reliability and safety, and the number of operation instruction words is as small as about a dozen words, it is easy to obtain a relatively high recognition rate. Speaker-type voice recognition devices have become mainstream.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Name of item to be corrected] 0005

[Correction method] Change

[Correction content]

Further, the recognition means of the voice recognition device generally compares the standard pattern data of a plurality of operation instruction words registered in advance with the pattern data of the operation instruction voice inputted from the microphone, and makes the pattern data similar to each other. It is known that a degree point is calculated using a similarity degree method such as a simple similarity degree method, and operation instruction data corresponding to an operation instruction word having pattern data with the highest degree of similarity (similarity point) is output. There is. However, the utterance sound of the operation instruction word uttered by the user is not always constant, and even if the utterance sound is made by the same utterer, the utterance sound of the same utterance sound is standard because of the unstable factors such as physical condition and mood of the day. Compared to when the pattern data is registered, the voice characteristics of the uttered sound often change slightly. Therefore, the degree of similarity with the input standard pattern data becomes low, which may lead to a reduction in the recognition rate.

[Procedure 3]

[Document name to be amended] Statement

[Correction target item name] 0018

[Correction method] Change

[Correction content]

Next, when the operator utters a voice for operating instruction, for example, "zoom up", while the voice controller 11 is waiting for the input of the voice for recognition, this is input to the microphone 9 as an acoustic signal. , To the voice controller 11 via the microphone wire 10.
The input signal is amplified by the amplifier 12 and then input to the voice recognition device 13, frequency-analyzed by the 8-channel BPF (not shown), and then converted into digital data by the A / D converter (not shown). It ─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] November 5, 1992

[Procedure Amendment 1]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 1

[Correction method] Change

[Correction content]

[Figure 1]

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Takushi Saito 2-43-2 Hatagaya, Shibuya-ku, Tokyo Inside Olympus Optical Co., Ltd.

Claims (1)

[Claims] 1. A surgical microscope having one or more drive means, a voice controller having a voice recognition means provided in a surgical microscope, and the drive means being operated by a voice. Is provided with means for calculating the standard pattern data of each operation instruction word of the voice recognition means and the pattern data of the input voice which are registered in advance, and updating the contents of the standard pattern data based on the calculation result. A voice controller for a surgical microscope.