JP2017009956A - Data recording device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data recording device for facilitating a work to convert a voice into text and the management of data.SOLUTION: A data recording device 10 includes: a storage part 30 in which voiceprint data for reference of a plurality of persons are stored: a microphone part 50 for detecting a voice; an identification part 80 for collating the voiceprint of the voice detected by the microphone part 50 with the voiceprint data for reference of a plurality of persons to identify a coincident speaker; and a text conversion part 90 for converting the voice detected by the microphone part 50 into text, therein text data converted into text are stored so as to be associated with the speaker identified by the identification part 80 in the storage part 30.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a data recording apparatus for recording voice as text.

  In hospitals and other medical settings, medical staff such as doctors and nurses can listen to multiple patients, record and take notes, and then convert the content heard into electronic data and record it in an electronic medical record. Has been done.

  By the way, in recent years, since technology for converting speech into text has been developed, it is possible to convert the heard speech into text by speech processing by a computer or the like (see, for example, Patent Document 1).

JP 2000-123007 A

However, for example, when interviews are performed continuously from a plurality of patients, it is necessary to record the patient's name as a memo or the like. Therefore, it takes time and effort to listen, and it is necessary to input the contents heard while referring to the memo to an electronic medical record or the like, which is complicated.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a data recording apparatus that makes it easy to convert voice into text and manage data.

  A data recording apparatus according to the present invention that achieves the above object includes a storage unit that stores voiceprint data for reference of a plurality of people, a microphone unit that detects voice, and a plurality of voiceprints of voice detected by the microphone unit. An identification unit for identifying a matching speaker by collating with the reference voiceprint data of a person, and a text-forming unit for converting the voice detected by the microphone unit into text, the identification unit identifying The text data converted into text is stored in the storage unit in association with the selected speaker.

  The data recording apparatus configured as described above compares the voice voice print detected by the microphone unit with the reference voice print data stored in the storage unit, identifies the speaker, and converts the text data into text. Since it can be stored in association with a person, it is easy to manage voice management and data.

It is a block diagram which shows the data recording device which concerns on embodiment. It is a top view which shows the data recording device which concerns on embodiment. It is a front view which expands and shows a data recording device. It is a side view which expands and shows a data recording device. It is a rear view which expands and shows a data recording device. It is a front view which shows the state which connected the data recording device to the computer with the cable. It is a flowchart which shows the flow of control of the data recording device which concerns on embodiment. It is the schematic for demonstrating the usage example of a data recording device. It is an expansion schematic for demonstrating the usage example of a data recording device. It is a side view which shows the modification of a data recording device.

  Embodiments of the present invention will be described below with reference to the drawings. In addition, the dimension ratio of drawing is exaggerated on account of description, and may differ from an actual ratio.

  A data recording apparatus 10 according to an embodiment of the present invention is an apparatus that converts voice data into text data and records it while automatically identifying a speaker. The data recording device 10 can be used when, for example, a medical worker such as a doctor or a nurse listens to a patient with presbycusis, converts voice data into text data while automatically identifying a speaker, Data input to the electronic medical record can be facilitated.

  As shown in FIGS. 1 to 5, the data recording device 10 selects a speaker from a voice print of a collected voice, a storage unit 30 in which a voice print or the like is stored, a control unit 40, a microphone unit 50 that collects sound, and the like. An identification unit 80 for identifying and a text unit 90 for converting voice data into text data are provided. The data recording apparatus 10 further performs various operations with the audio processing unit 60 that processes audio, the bone conduction speaker unit 70 that transmits sound to the patient, and the bone conduction microphone unit 100 that collects sound from the bone of the patient. The operation part 110 for this, the 1st switch part 120, the connection part 140 which can connect the connector 311 of the communication cable 310, the housing | casing 20 which accommodates these, and the strap 130 are provided.

  The casing 20 includes a contact surface 23 on which the bone conduction speaker unit 70 and the bone conduction microphone unit 100 are disposed and pressed against the skin, and an operation surface 24 opposite to the contact surface 23 and on which the operation unit 110 and the like are disposed. Is formed. On the side surface between the contact surface 23 and the operation surface 24, a concave gripping portion 21 and an annular connection portion 22 to which the strap 130 is connected are formed so that the operator can easily grip the surface.

  The storage unit 30 stores in advance, as reference voice print data, an operation program executed by the control unit 40 and voice print data of a plurality of patients or medical workers. A voiceprint is a graph in which a voice is subjected to frequency analysis, the vertical axis represents frequency, the horizontal axis represents time, and the intensity change of the spectrum of the speech waveform is displayed in contour lines. Voiceprints have different shapes depending on the person. In addition, the storage unit 30 stores various parameters used in the voice processing unit 60, the identification unit 80, the text unit 90, and the like. The storage unit 30 further stores digital voice data digitized after being collected by the microphone unit 50, text data converted to text by the text unit 90, and the like.

  The microphone unit 50 is a microphone that collects voices uttered by the operator and the patient, generates a voice signal from the collected voices, and transmits the voice signal to the voice processing unit 60.

  The audio processing unit 60 has a function of converting an audio signal input from the microphone unit 50 into digital audio data in a format such as MP3 or WAV. Furthermore, the audio processing unit 60 has a function of performing frequency conversion and volume adjustment on the audio signal input from the microphone unit 50 for each frequency band stored in the storage unit 30.

  The audible range in which a human ear can be heard is about 20 to 20000 Hz, but the frequency range necessary for listening to human speech is about 125 to 8000 Hz. Among these, in senile deafness, it is difficult to hear high frequencies, and in particular, it is difficult to hear from a high frequency part of 4000 Hz or higher.

  By the way, although there are individual differences, as an example, many vowels are included in the low band of 125 to 1500 Hz, voiced consonants are included in the middle band of 1500 to 3000 Hz, and unvoiced consonants are included in the high band of 3000 to 6000 Hz. It is. Therefore, in presbycusis, the low band containing vowels is easier to hear, the middle band containing voiced consonants is harder to hear than the low band, and the high band containing unvoiced consonants is more difficult to hear. In senile deafness, vowels can be heard even at about 30 dB, but voiced consonants are difficult to hear unless they are about 40 dB or more, and unvoiced consonants are difficult to hear unless they are about 50 dB or more. For this reason, the audio processing unit 60 can perform frequency conversion and volume adjustment for each band. For example, the voice processing unit 60 does not perform frequency conversion for a low band (125 to 1500 Hz) where vowels exist, and amplifies it to 30 dB or more when it is less than 30 dB. And the audio | voice processing part 60 frequency-converts the middle band (1500-3000 Hz) with many voiced consonants to 1200 Hz or less, and when it is less than 40 dB, it amplifies to 40 dB or more. The high band (3000 to 6000 Hz) with many unvoiced consonants is frequency-converted to 3000 Hz or less that is easy to hear even for presbycusis and is amplified to 50 dB or more when it is less than 50 dB. Thus, vowels, voiced consonants, and unvoiced consonants can be modulated and amplified for each band, and almost all speech can be converted to be recognizable. Note that the audio processing unit 60 may not perform audio frequency conversion or volume adjustment.

  The audio processing unit 60 transmits the frequency-converted and amplified audio signal to the bone conduction speaker unit 70.

  The bone conduction speaker unit 70 is a speaker including a bone conduction contact 71 and an actuator 72 that generates vibration. The actuator 72 generates vibration by the signal received from the sound processing unit 60 and vibrates the contact 71. The contact 71 is exposed from the housing 20 and can contact the patient. The contact 71 comes into contact with the patient's skin 202 and transmits vibrations to the skull 201 (see FIG. 9) located directly under the skin 202 through the skin 202. The bone conduction speaker unit 70 can be switched on and off by an operation in the operation unit 110. As a result, the bone conduction speaker unit 70 can be used only when a conversation cannot be established without using the bone conduction speaker unit 70.

  The bone conduction microphone unit 100 includes a highly sensitive vibration sensor element 101 that measures vibration generated in a patient's vocal cords via the skull 201, generates a voice signal from the collected vibration, and transmits the voice signal to the voice processing unit 60. . The bone conduction microphone unit 100 can be switched on and off by an operation in the operation unit 110. As a result, the bone conduction microphone unit 100 can be used only when a conversation cannot be established without using the bone conduction microphone unit 100.

  The identification unit 80 converts voice data obtained by the microphone unit 50 and the bone conduction microphone unit 100 into voiceprint data. Next, the identification unit 80 collates the voiceprint data converted from the voice data with a plurality of reference voiceprint data stored in the storage unit 30, and matches the reference voiceprint data that matches with a known speaker identification technique. To identify the speaker.

  The text unit 90 converts the voice data into text data. The text unit 90 can convert the numbers converted into characters into numerical values. In addition, the text converting unit 90 can recognize the numerical value as body temperature from characters such as “degree”, “minute”, “temperature” etc. before and after the numerical value. In the case of body temperature, a numerical value is recorded as body temperature in another segment for storing body temperature data. The text converting unit 90 inserts a line feed into the text data when there is no voice input to the microphone unit 50 for 2 seconds or more. The voice data and the text data are stored in the storage unit 30 in association with a person (for example, a medical worker and a patient) specified by a voiceprint and a time (for example, a date and time when recording is started). Digital audio data and text data can be associated by using, for example, a reference voice print data name and a recording start time as a file name. Note that the types and order of items used for these file names (for example, reference voice print data name, recording start time, etc.) can be arbitrarily set and changed. In addition, the order of these file names can be arbitrarily set and changed. It can also be automatically arranged in order of recording time.

  The operation unit 110 is a part that is arranged on the operation surface 24 of the housing 20 and performs various operations. The operation unit 110 includes an image monitor 111, a second switch unit 112 for the bone conduction speaker unit 70, and the bone conduction microphone unit 100. And a third switch section 113 for use. The image monitor 111 can display the name of the speaker identified by the identifying unit 80, or can display a text that is converted from speech into text. The image monitor 111 is formed with a display device such as an LCD (Liquid Crystal Display) or an organic EL (Electro-Luminescence) display. The second switch unit 112 is a switch for switching the bone conduction speaker unit 70 on and off. The third switch unit 113 is a switch for switching the bone conduction microphone unit 100 on and off.

  As shown in FIGS. 4 and 6, the connection unit 140 is a connector that can connect the connector 311 of the communication cable 310 from the outside in order to transmit and receive data to and from the storage unit 30. The communication cable 310 is, for example, a USB cable, and the recorded data can be transferred to the electronic medical record by connecting to the computer 300 that can access the electronic medical record through the communication cable 310. The connection unit 140 also serves to charge a battery (not shown) in the housing 20 via the communication cable 310.

  In addition, as a method of transmitting and receiving data, it is possible to perform wirelessly using a communication technique such as NFC (Near Field Communication) instead of using a cable.

  1-4, the 1st switch part 120 is provided in the housing | casing 20 so that operation | movement of the data recording device 10 may be started, when the contactor 71 is made to contact the skin 202. As shown in FIGS. The first switch part 120 includes, for example, a convex part 121 that is biased in a state of projecting by a spring (not shown), and the convex part 121 projects by being pushed in contact with the skin 202. Two contact points (not shown) which are located apart from each other. When the two contacts come into electrical contact, a current flows and the operation of the data recording device 10 is started. Then, when the convex portion 121 is separated from the skin 202, the convex portion 121 is projected by the spring, the two contact points are separated, and the operation of the data recording apparatus 10 is stopped. The configuration of the first switch unit 120 is not limited to this. For example, the operation of the data recording apparatus 10 is started when two electrodes are in contact with the skin 202 and a current flows through the skin 202. Alternatively, it may be a temperature sensor that starts the operation of the data recording device 10 when the temperature rises in contact with the skin 202.

  The strap 130 is a string-like flexible member attached to the housing 20 so that the operator can be hung on the neck or the like. The operator can carry the strap 130 in a state where the strap 130 is hung on the neck or the like, and can use the data recording apparatus 10 by gripping the grip portion 21 while the strap 130 is hung on the neck.

  The control unit 40 includes a CPU (Central Processing Unit) and an operation program, and includes a microphone unit 50, a voice processing unit 60, a bone conduction speaker unit 70, a bone conduction microphone unit 100, an identification unit 80, a text unit 90, and an operation unit. The operation of the 110 and the first switch unit 120 is comprehensively controlled. When one of the associated text data and voice data is deleted, the control unit 40 also deletes the other. Data can be erased from the image monitor 111 which is a touch panel of the operation unit 110, for example.

  Next, a method of using the data recording apparatus 10 according to the present embodiment will be described with reference to the flowchart shown in FIG. 7, taking as an example a case where an operator such as a doctor or a nurse listens to a patient with senile deafness. To do. Here, a case where the bone conduction speaker unit 70 and the bone conduction microphone unit 100 are used will be described as an example.

  First, the operator turns on (operates) the second switch unit 112 and the third switch unit 113 with the strap 130 on the neck. As a result, the bone conduction speaker unit 70 and the bone conduction microphone unit 100 can be operated.

  Next, as shown in FIGS. 8 and 9, the operator grasps the grasping portion 21 and brings the contact 71 and the bone conduction microphone portion 100 into contact with the skin 202 covering the skull 201 of the patient.

  When the bone conduction microphone part 100 and the contact 71 are brought into contact with the skin 202, the convex part 121 of the first switch part 120 is pushed in, and the two contacts inside the first switch part 120 come into contact with each other. The operation of the data recording device 10 is started (step S1).

  When the operation of the data recording device 10 is started, a conversation is collected by the microphone unit 50 and the bone conduction microphone unit 100. When the voices of the operator and the patient are input to the microphone unit 50 or the bone conduction microphone unit 100, the voice data is transmitted to the voice processing unit (step S2). Next, the identification unit 80 converts voice data obtained by the microphone unit 50 and the bone conduction microphone unit 100 into voiceprint data, and the voiceprint data is converted into a plurality of reference voiceprint data stored in the storage unit 30. The speaker is identified by collating and identifying the matching reference voiceprint data (step S3). The identified speakers are, for example, medical personnel and patients. The identified speaker is displayed on the image monitor 111 (step S4). If the identified speaker is wrong, the first switch 120 can be moved away from the skin and pressed again to resume.

  The audio processing unit 60 converts the frequency of the received audio data according to the frequency band (step S5) and adjusts the volume according to the frequency band (step S6). As a result, high-frequency sound that is difficult to hear is converted to low frequency, and high-frequency sound that is difficult to hear is amplified, so that it is easy to hear even if it is hard to hear.

  Next, the sound processing unit 60 transmits the sound signal amplified by converting the frequency to the bone conduction speaker unit 70. In the bone conduction speaker unit 70, the contact 71 is vibrated by the actuator 72, and the skull 201 located immediately below the skin 202 is vibrated (step S7). Since vibration is transmitted to the cochlea 200 via the skull 201, even if the functions of the outer ear, the middle ear, and the inner ear are deteriorated, it is easy for the patient to recognize high sounds and sounds that are difficult to hear.

  The voice data collected by the microphone unit 50 and the bone conduction microphone unit 100 is converted into text data by the text unit 90 (step S8). The text forming unit 90 determines whether the character includes a number (step S9), and when determining that the character includes a number, converts the character into a numerical value (step S10). Further, it is determined whether the numerical value is body temperature data (step S11). If it is determined that the temperature is body temperature, the numerical value is recorded as a body temperature in another segment that stores the body temperature data (step S12). Further, the text conversion unit 90 determines whether there is no voice input to the microphone unit 50 for 2 seconds or more (step S13), and if there is no voice input for 2 seconds or more, inserts a line feed into the text data (step S13). S14).

The text data converted into text in the text converting unit 90 is displayed on the image monitor 111 (step S15). Thereby, a conversion mistake etc. can be confirmed. For example, if the image monitor 111 is a touch panel, a conversion error or the like can be corrected.

  After the conversation is finished, the operator separates the data recording device 10 from the skin 202 while holding the grip portion 21. Thereby, the 1st switch part 120 turns off, and the input / output of the audio | voice by the microphone part 50, the bone conduction speaker part 70, and the bone conduction microphone part 100 is complete | finished (step S16).

  Thereafter, the voice data is converted into digital voice data by the voice processing unit 60, and the digital voice data and the text data are identified by the voiceprint (for example, a healthcare worker and a patient) and the date (for example, recorded). The date and time are started and stored in the storage unit 30 (step S17).

  As an example, when the operator is identified as “A010”, the patient is “B020”, and the recording start date is “10:10 on May 1, 2015”. The file name of the digital audio data can be “A01B020 — 201505011010.mp3”, and the file name of the text data can be “A01B020 — 201015011010.txt”. After the data is stored, the operation of the data recording device 10 ends. Note that the types and order of items used in these file names (for example, a symbol representing an operator, a symbol representing a patient, a recording start time, etc.) can be arbitrarily set and changed. In addition, the order of these file names can be arbitrarily set and changed. It can also be automatically arranged in order of recording time.

  The operator can sequentially perform the above-described recording for a plurality of patients. After the recording of a plurality of persons is completed, the operator connects the cable 310 to the connection unit 140 and connects to the computer 300 that can be connected to the electronic medical chart as shown in FIG. Thereby, the recorded digital audio data and text data can be transferred to the electronic medical record. Since the transferred digital audio data and text data are associated with each other by the operator, patient and date / time data, the management becomes easy. In addition, the data recording device 10 can be charged by connecting the cable 310 to the connection unit 140.

  As described above, the data recording apparatus 10 according to the present embodiment includes the storage unit 30 that stores voiceprint data for reference of a plurality of people, the microphone unit 50 that detects audio, and the audio detected by the microphone unit 50. Data recording apparatus comprising: an identification unit 80 for identifying a matching speaker by comparing the voiceprint of the voice with a plurality of reference voiceprint data, and a text unit 90 for converting the voice detected by the microphone unit 50 into text 10. Since the data recording apparatus 10 configured as described above can identify the speaker by comparing the voiceprint detected by the microphone unit 50 with the reference voiceprint data stored in the storage unit 30, It can be made into text efficiently while being related to.

  In addition, since the text data that has been converted into text is stored in the storage unit 30 in association with the speaker identified by the identification unit 80, it is not necessary to take notes, and the operation can be performed easily and accurately. Management of text data becomes easy.

  The data recording apparatus 10 further includes a voice processing unit that digitizes the voice detected by the microphone unit 50, and the storage unit 30 stores the digital voice data digitized together with the text data. Data and digital audio data can be easily managed.

  Further, since the text data and the digital voice data have information on the time when the recording was performed, the text data and the digital voice data can be managed in association with each other.

  Moreover, since the text conversion part 90 converts the number which is a text into a numerical value, management of numerical values, such as body temperature, becomes easy, for example.

  In addition, since the text conversion unit 90 breaks the text when the voice detected by the microphone unit 50 does not exist for a predetermined time, the text visibility and handling are improved.

  In addition, since the data recording device 10 includes the bone conduction speaker unit 70 that outputs the sound detected by the microphone unit 50 as vibration, the data recording device 10 can effectively transmit the sound to a person with hearing loss. Can be recorded well.

  Moreover, since the data recording apparatus 10 has the bone conduction microphone part 100 in which the microphone part 50 detects a sound from the vibration of the part which contacts, even if it is a voice of a person with a difficult conversation, it is effective via a bone. Therefore, it is possible to record data well.

  Note that the present invention is not limited to the above-described embodiments, and various modifications can be made by those skilled in the art within the technical idea of the present invention. For example, in the data recording device 150 which is a modified example shown in FIG. 10, the strap 160 includes a wire 162 provided with a clip 161 on the side opposite to the side to which the housing 170 is connected. A reel mechanism 163 that can be wound may be provided. The reel mechanism 163 is, for example, a spring type. Further, the reel mechanism 163 includes a switch mechanism 180 that starts the operation of the data recording device 150 when the wire 162 is pulled out by a predetermined length or more. The switch mechanism 180 is, for example, an optical sensor that detects the winding amount (thickness) of the wire 162 wound around the roll of the reel mechanism 163. The data recording device 150 is carried in a state where the wire 162 is wound around the reel mechanism 163 and the clip 161 is attached to the clothes of the operator. When the operator brings the casing 170 close to the patient and the wire 162 is pulled out from the reel mechanism 163 by a predetermined length or longer, the operation of the data recording device 150 is started. In addition, the same code | symbol is attached | subjected to the site | part which has the same function as the above-mentioned embodiment, and description is abbreviate | omitted.

  Further, the data recording apparatus may not include a bone conduction speaker unit or a bone conduction microphone unit.

  Further, the data recording apparatus may be configured by a portable terminal such as a smartphone or a tablet terminal and an application (program).

  Further, the clip 161 may be configured by changing to a phone plug (φ3.5 mm mini plug). In this case, the data recording apparatus can be suitably used for portable terminals such as smartphones and tablet terminals and applications (programs), and the configuration of the data recording apparatus 150 can be simplified.

10, 150 data recording device,
30 storage unit,
40 control unit,
50 Microphone part,
60 voice processing unit,
70 bone conduction speaker,
80 identification part,
90 Texting part,
100 bone conduction microphone,
110 operation unit,
111 image monitor,
200 Cochlea,
201 skull,
202 Skin.

Claims (7)

A storage unit for storing voiceprint data for a plurality of people;
A microphone for detecting voice;
An identification unit for identifying a matching speaker by collating voice voice prints detected by the microphone unit with a plurality of reference voice print data;
A data recording unit that stores the text data that has been converted into text in association with the speaker identified by the identification unit; apparatus. An audio processing unit for digitizing audio detected by the microphone unit;
The data storage device according to claim 1, wherein the storage unit stores digital audio data digitized together with the text data.   The data recording apparatus according to claim 2, wherein the text data and the digital audio data have information on a time at which recording was performed.   The data recording device according to claim 1, wherein the text converting unit converts a number that is text into a numerical value.   The data recording device according to any one of claims 1 to 4, wherein the text converting unit changes the text to a new line when the sound detected by the microphone unit does not exist for a predetermined time.   The data recording apparatus according to any one of claims 1 to 5, further comprising a bone conduction speaker unit that outputs sound detected by the microphone unit as vibration.   The data recording apparatus according to any one of claims 1 to 6, wherein the microphone unit includes a bone conduction microphone unit that detects sound from vibration of a contacted part.

Images