Input support system and input support system for medical record

Abstract

To provide a support system configured to easily prepare a summary from text data which is obtained from input voice, and to input patient information, such as nursing information, to storage means, such as an electronic medical chart, or the like.SOLUTION: An input support system includes a terminal device 1, voice recognition means 3 which recognizes input voice to be converted into text data, and processing means 4. The terminal device includes input means for voice information, storage means, and display means. The display means is partitioned into a first display region for displaying text data, and a second display region for displaying information obtained by extracting only desired elements from the text data. The processing means causes the whole of the text data to be displayed in the first display region, and extracts item target information related to predetermined extraction items from the text data, to be displayed in the second display region. An input support system for medical record is configured to allow a terminal device to access an electronic medical chart.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a system for assisting in inputting information into a storage means after correcting voice data input by voice, and to an input support system that can be used to assist in input into a medical record system such as an electronic medical record system. [Background technology]

[0002] A system has been developed for correcting character strings when the character strings are converted from voice data into text and stored in a specific storage means (see Patent Document 1). This system uses a mobile terminal, and edits the converted character strings by searching for character strings related to a specific word, displaying them as candidates for deletion, and deleting the relevant character strings. However, when it is necessary to correct the converted character strings in addition to editing the character strings to be deleted, a more user-friendly system is desired.

[0003] As a result, a system that allows correction of character strings on the display screen of a mobile terminal has been developed (see Patent Document 2). This system configures multiple display areas on the display screen of a mobile terminal, and in one of the display areas, words that are divided into word units and that make up a word dictionary for speech recognition are displayed, and the display area that displays correction candidate words is used to correct the character string word by word. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2016-029466 A [Patent Document 2] JP 2018-180883 A [Patent Document 3] JP 2001-344346 A [Patent Document 4] JP 2007-293600 A Summary of the Invention [Problem to be solved by the invention]

[0005] The prior art disclosed in the above-mentioned Patent Documents 1 and 2 are both technologies for deleting or modifying part of the entire character string obtained as a result of recognizing input speech, and were not systems having the function of extracting specific words, character strings, sentences, etc. from the text-converted character string, aggregating the extracted elements, or summarizing them as a summary.

[0006] On the other hand, devices that enable voice input have been developed as input support for medical management records such as electronic medical records (see Patent Documents 3 and 4). The technology disclosed in Patent Document 3 is a device for editing and inputting information to be stored in an electronic medical record by voice, and changes the input method from manual input to voice input. In addition, the technology disclosed in Patent Document 4 displays candidates for phrase breaks and homonyms when converting to text, and allows editing of character strings by referring to these.

[0007] However, the information to be stored in the electronic medical record includes, for example, daily vital sign information or observation information for hospitalized patients, as well as various other information such as progress observation information. However, this information was manually entered by nurses individually as nursing information. In addition, when operating a terminal for each vital sign, multiple measuring devices could not be used at the same time, which could result in patients having to wait. For this reason, the method of having nurses take notes of information such as patients' vital signs and then inputting them separately into the electronic medical record all at once was adopted. This meant that the daily task of entering nursing information took time, increasing the burden on nurses. The same was true when doctors made rounds.

[0008] Furthermore, the content to be recorded as nursing information is often not all of the information entered by voice, but only the key information from the electronic medical record, such as vital signs, and there is a demand for extracting only the key information from the information entered by voice, but such technology has not yet been developed.

[0009] The present invention has been made in consideration of the above-mentioned points, and its object is to provide a support system which can easily extract and consolidate specific character strings, etc. from data obtained by converting input voice into text, enables the correction of the consolidated information, and enables patient information such as nursing information to be input into a storage means such as an electronic medical record. [Means for solving the problem]

[0010] Therefore, the present invention relates to an input support system, which supports input to a predetermined storage means, and includes a terminal device capable of transmitting and receiving, a voice recognition means for recognizing input voice and converting it into text data, and a processing means for processing the text data recognized by the voice recognition means, and is characterized in that the terminal device includes a means for inputting voice information, a storage means for storing the voice data and the text data input by the input means, and a display means for displaying the text data, and the display means is divided into a first display area for displaying the text data and a second display area for displaying information obtained by extracting only desired elements from the text data, and the processing means displays the entire text data in the first display area, and extracts item target information related to predetermined extraction items from the text data and displays it in the second display area, and when a character string displayed in either one of the first or second display areas of the terminal device is corrected, the processing means processes the input support system so as to correct the same character string displayed in the other display area.

[0011] According to the above configuration, the input voice is converted into text data and displayed in the first display area of ​​the display means, and item target information extracted from the text data can be displayed in the second display area along with a predetermined extracted item. The content displayed in the second display area is a collection of main information to be recorded, and by saving this, an extract-type summary of the recorded information can be created. In addition, when the input information is patient information such as nursing information, it can be used as a support tool that allows nursing information to be saved to be input to a saving means such as an electronic medical record using extracted information. Note that the character strings displayed in the first and second display areas are all or part of the text data stored in the storage means, so that when a part of the character strings displayed on one of the display means is corrected, the text data stored in the storage means can be updated, and the same content can be corrected for the character strings displayed in the other display area based on the updated text data.

[0012] In the above configuration, the processing means can be configured to assign an ordinal number to each character of a string constituting the text data, store each character together with the ordinal number in the storage means, display in the first display area a string in which each character is arranged in the order of the ordinal numbers stored in the storage means, and display in the second display area a string in which each character of the item target information is arranged based on the order of the ordinal numbers assigned to the character, and when some characters displayed in either the first or second display area are corrected, update and re-assign the ordinal number to the entire string after correction, thereby updating the information stored in the storage means.

[0013] According to the above configuration, the character string converted into text by the voice recognition is stored in the storage means together with the ordinal number, so that the entire text data can be displayed by displaying the character string stored in the storage means in accordance with the ordinal number. This is the information displayed in the first display area. Also, in the second display area, the item target information related to the extracted item is displayed based on the order of the ordinal numbers of the character string, so that only the desired information can be displayed. When correcting the character string displayed in the first and second display areas, the ordinal number is updated and reassigned to the entire corrected character string, so that the character string displayed in both display areas can be corrected. The corrected character and ordinal number are updated and stored in the storage means, so that the latest information can be displayed in the first and second display areas by referring to the updated ordinal number.

[0014] In the invention having the above configuration, the voice recognition means is provided in an external device capable of transmitting and receiving from the terminal device, and the terminal device can temporarily store text data received from the external device in the memory means.

[0015] According to the above configuration, it is not necessary to provide a voice recognition unit for each terminal device to be used, and it is possible to eliminate delays in information processing in the terminal device. In this case, the external device may be, for example, a server used as a storage unit, or a voice recognition unit available on the web may be used.

[0016] In the invention having the above configuration, when the processing means detects a plurality of pieces of item target information, the processing means may sequentially display all of the plurality of pieces of item target information in the second display area.

[0017] According to the above configuration, since an erroneous utterance (such as a slip of the tongue) may occur during voice input, it is conceivable that the utterance itself may be corrected immediately after the utterance. When processing text data based on such input voice, multiple pieces of target item information related to the same extraction item are detected, so all of these pieces of target item information are displayed in sequence in the second display area, and an operation for deleting the erroneous target item information allows only a specific piece of target item information to be selected in the end. In this case, even if incorrect information is input (a slip of the tongue) during voice input, the correct information can be selected in the second display area by subsequently inputting (repeat) the correct information, thereby eliminating the need to individually correct the character strings displayed on the display means.

[0018] On the other hand, the present invention relating to an input support system for medical records uses any of the inventions relating to the input support systems described above, and is characterized in that the storage means stores examination information, treatment information and other medical management information for each patient, the terminal device is able to access the medical information for each patient stored in the storage means, and the processing means processes items indicating medical elements as the extracted items and information to be recorded as the medical information for each patient as the item target information, respectively.

[0019] According to the above configuration, it can be used as a support system for inputting key information into medical records such as electronic medical charts. That is, by inputting the extracted target item information displayed in the second display area from the voice-input information together with the extracted items, the key contents of the medical management information for each patient stored in the electronic medical chart, etc. can be additionally recorded. The input means in this case is voice input, and the character string converted into text by voice recognition can be corrected by correcting the character string displayed in the first or second display area of ​​the display means.

[0020] In the invention of the above configuration, the storage means can be configured to be capable of storing the text data displayed in the first display area and the item target information displayed in the second display area, and the processing means can be configured to distinguish between the text data and the item target information and process to send either one or both of them to the storage means.

[0021] According to the above configuration, it is possible to distinguish between information corrected for the entire character string obtained by converting the input voice into text by voice recognition, and information extracted from the entire corrected character string, and to save either one or both for the medical record. Information collected exclusively as main information (information displayed in the second display area) should be saved as nursing information, etc., but patient-specific information that is not directly related to nursing (non-main) and is included in the voice input information can be managed as part of the nursing record by saving the entire character string (information displayed in the first display area). Effect of the Invention

[0022] According to the present invention, since it is possible to display the entire character string obtained by converting the input voice into text, and to display only the target item information to be extracted from the entire character string, it is possible to easily consolidate the main information based on the input data and provide it for input for storage. Furthermore, since the target item information is information related to predetermined extraction items, if a character string obtained by converting the input data into text is obtained, the target item information contained in these character strings will necessarily be reflected as the content of the extracted item, and it is possible to realize assistance in inputting the main information.

[0023] In addition, since patient information such as nursing information can be selectively sent to a storage means such as an electronic medical record as either the entire string or the aggregated information, or both, it is possible to assist in the cumulative input of medical information stored in an electronic medical record, etc. [Brief description of the drawings]

[0024] [Figure 1] FIG. 2 is an explanatory diagram showing a display example of a terminal device in the embodiment. [Diagram 2] FIG. 1 is an explanatory diagram showing an overview of an input support system. [Diagram 3] FIG. 2 is an explanatory diagram showing a configuration example of a display format in the embodiment. [Figure 4] FIG. 2 is an explanatory diagram showing a configuration example of a display format in the embodiment. [Diagram 5] FIG. 4 is an explanatory diagram showing a processing flow up to displaying a voice recognition result in the embodiment. [Figure 6] FIG. 2 is an explanatory diagram showing a processing flow for structuring in the embodiment. [Figure 7] FIG. 11 is an explanatory diagram showing a processing flow for extracting target item information in the embodiment. [Figure 8] FIG. 13 is an explanatory diagram showing a modified example of the embodiment. [Figure 9] FIG. 11 is an explanatory diagram showing a processing flow in a modified example of the embodiment. [Figure 10] FIG. 11 is an explanatory diagram showing a processing flow for extracting target item information in a modified example of the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] <Overview of the embodiment> Hereinafter, an embodiment of the present invention will be described with reference to the drawings. An outline of this embodiment is shown in Figs. 1 and 2. Fig. 1 shows an outline of a terminal device 1 that consolidates and displays text data (text data) that is input voice data and information extracted from the text data, and Fig. 2 shows the overall configuration of an input support system. In addition, since the present invention relates to an input support system, it is possible to consolidate information extracted uniformly by selecting in advance any extraction items from various text data. For example, it may be minutes of a meeting or a conference for a specific purpose (with limited content), and it can be used to consolidate record items to be stored in an already constructed management system. In this embodiment, a case will be illustrated in which major information related to nursing that should be recorded in an electronic medical record is consolidated to be input information.

[0026] As shown in FIG. 1, in this embodiment, a screen 10 of a terminal device (a smartphone is shown as an example of a mobile terminal) 1 is divided into two, and the first display area 11 displays the entire text data, and the second display area 12 displays main information obtained by extracting desired items from the text data while aggregating them. The first display area 11 is provided with a display field 21 with a heading of "input sentence", which uses voice data previously input by voice and displays the result of converting the voice data into text by voice recognition. On the other hand, the second display area 12 is provided with a heading of "item result", and further includes target item information fields 22a to 22h in which extracted items are listed and target item information is displayed for each extracted item. In the target item information fields 22a to 22h, information related to the extracted items is extracted from the text data and displayed individually. The extracted items and target item information displayed in the second display area 12 constitute the entire main information.

[0027] <Embodiments Related to Input Support System> As shown in Fig. 2(a), the system for displaying text data and summaries on the terminal device 1 is constructed so that the terminal device 1 can access an API (Application Programming Interface) 3, a structuring server 4, and an electronic medical record system 5 via a network 2. The API 3 used here is an API for using streaming speech recognition from a browser, and for example, the Web Speech API can be used. The Web Speech API is an API established by the World Wide Web Consortium (W3C), and can be incorporated using JavaScript.

[0028] Here, "structuring" is a term meaning to divide text data (text data) into extracted items, and specifically, to extract character strings related to the extracted items as target item information and display them in the second display area 12. The structuring server 4 is a server that stores a program for the structuring, and corresponds to the processing means of the present invention. The structuring server 4 is constructed so as to be accessible from the terminal device 1, but the program for the structuring may be constructed so as to be stored in the terminal device 1. When the program is stored in the terminal device 1, the terminal device 1 also functions as the processing means. For example, it can be constructed on a Web browser, and there is a method of accessing a specific URL or downloading a script to the terminal device 1 and processing it on the browser. For example, Java script can be used as this type of processing method. When constructed on a Web browser, the configuration of the first and second display areas 11 and 12 can also be processed as a screen configuration. In this case, it can also be constructed using Java script, for example.

[0029] The electronic medical record 5 is assumed to be accessible to the terminal device 1, and in particular, an area is provided for inputting nursing information and recording it in the electronic medical record, and data converted into text by manual input (keyboard input) or the like can be input arbitrarily. In this embodiment, it is assumed that the system can record data converted into text in a predetermined area.

[0030] Note that the Web Speech API and electronic medical record system are not the main part of this invention, and since already developed technologies can be used for them, they will not be described in detail, but rather the explanation will focus on the content specific to the input support system.

[0031] The support system constructed as described above is operated by the user's operation of the terminal device 1, as shown in FIG. 2(b). That is, the user inputs and stores voice data using the voice input device (input means) and the recording function (storage means) of the terminal device 1, transmits the data to the outside, and executes the voice recognition process by the Web Speech API 3. The terminal device 1 receives the result (text data) of the voice recognition by the API 3, and displays the result on the terminal device 1 (first display area 11). Furthermore, the recognition result is transmitted to the structuring server 4 for structuring process (or by a stored program), and the target item information is classified for each extraction item, and the result is displayed on the terminal device 1 (second display area 12). The user can confirm the contents displayed on the terminal device 1 (first and second display areas 11, 12), correct them as necessary, and then instruct the electronic medical record 5 to transmit the confirmed contents, so that the electronic medical record 5 can be made (reflected) as aggregated information (main information).

[0032] It should be noted that the instructions for individual data transmission may be configured to be processed sequentially when the voice input is completed, except for the final transmission of the consolidated information to the electronic medical record 5. In addition, the transmission to the electronic medical record 5 may be configured to transmit not only the consolidated information, but also the contents of the corrected recognition results (entire text data) and to store them in the electronic medical record 5 system.

[0033] Here, this embodiment is configured as a system capable of simultaneously correcting both information, such as by recording the aggregated state of the speech recognition result (text data) and the extracted information (structuring result) in the electronic medical record 5. Therefore, the information extracted from the text data (target item information) is not constructed separately from the text data, but is displayed in the first and second display areas 11, 12 with the same data displayed, but only the display format changed. In order to construct such a notation format, in this embodiment, an ordinal number is assigned to each character to be displayed as the speech recognition result (text data), and a common character is displayed by a common ordinal number.

[0034] Configuration examples of the above display format are shown in Figures 3 and 4. Note that ordinal numbers are assigned to character strings in the results of speech recognition (text data), and the converted character strings are stored in a storage means such as a memory together with the ordinal numbers, and the stored character strings are displayed in display areas 11 and 12 with reference to the ordinal numbers. Therefore, although ordinal numbers are not assigned to the character strings displayed in first display area 11, for the sake of explanation, an example of the state of ordinal numbers assigned using character strings displayed in display areas 11 and 12 is shown.

[0035] As shown in Figure 3, character strings in text data are sorted according to the input voice data as a result of voice recognition, and ordinal numbers are assigned consecutively from "0" to the end in this order. When using the aforementioned Java script, ordinal numbers can be assigned to character strings starting from "0" by default. In the example shown, "Mr. A" at the beginning of the sentence is assigned the numbers "0", "1", and "2", and "Masu" at the end of the sentence is assigned the numbers "80", "81", and "82".

[0036] Figure 4 shows the state in which related target item information for each extracted item is extracted from the text data described above and displayed in the second display area 12.As shown in Figure 4, the strings (target item information) extracted for each extracted item by structuring processing and displayed for each item display characters with the same ordinal numbers as the strings in the text data described above.

[0037] Therefore, while referring to a single text data (data stored in a memory means), characters with the same ordinal numbers assigned to character strings in the text data are simply displayed individually in both display areas 11, 12, and are not configured as independent, individual character strings.

[0038] Therefore, when correcting the character strings displayed in both display areas 11, 12 as described above, whether correcting the character string displayed in first display area 11 or correcting the character string displayed in second display area 12, the correction is made on a character basis that has the same ordinal number assigned, so that by correcting either one, the corrections to both can be completed at the same time.

[0039] Since the text data itself is modified for a character string after modification, by updating the assigned ordinal number, a new ordinal number is assigned to the modified character string, and the character string is displayed in both display areas 11, 12 according to the updated content. The ordinal number is updated every time a character is modified, so that unmodified character strings are not changed on display screens 11, 12. In particular, when considering the case where the number of characters is changed, it is clear that the ordinal number should be updated every time a modification is made.

[0040] <Processing flow> Next, a process flow when using the input support system described above will be explained. Fig. 5 shows the process flow from when the use of this system is started to when the recognition result is displayed. As shown in this figure, when the use of the input support system is started, the display screen is basically divided into a first display area and a second display area (S101). This display screen makes it possible to confirm that the system is up and running. This type of screen configuration can be constructed using Java script as described above.

[0041] Next, recording is started using the recording function of the terminal device. By inputting voice (S102), recording starts (S103) and continues until an operation to end the recording is performed (S104). When an operation to end recording is performed, recording ends (S105). Next, an operation to start voice recognition is performed, and it is determined whether or not such an operation has been performed (S106). If no operation has been performed, an operation to interrupt the process is confirmed (S107). If an operation to interrupt the process is performed, the process ends and the input voice is maintained in a state where it has been saved by the recording function of the terminal device. This is to function in the case where only voice input is performed and then voice recognition and other processes are performed all at once.

[0042] When an operation to start voice recognition is performed, the recorded voice data is processed and converted to text (S108). In this case, if an external Web Speech API is used for voice recognition, the voice data is transmitted. If the same program is stored in the terminal device, the processing is performed within the terminal device. When voice recognition is performed, the result (text data) is obtained and temporarily stored in the memory unit (S109). The contents of the stored text data are then displayed in the first display area (S110), allowing the user to check the converted text string.

[0043] The process flow for extracting information for each extraction item using character strings converted to text by speech recognition and aggregating this information is shown in Figures 6 and 7. The process here is exclusively structuring processing, and can be processed by connecting (sending) to the structuring server in the above-mentioned system, or can be processed within the terminal device by storing the program in the terminal device.

[0044] When aggregating information extracted for each extraction item (target item information), first assign ordinal numbers to character strings in the text data (S201), as shown in Fig. 6. If you use Java script as described above, you can assign ordinal numbers sequentially starting from "0" by default.

[0045] Next, the predefined extraction items are read out as items to be extracted (S202). The extracted items are ranked, and predetermined target item information is extracted based on words related to the extracted items from the first to the final order. Related words refer to character strings set as words for extracting desired information, such as the same word as the extracted item, as well as words having the same meaning such as synonyms and similar words. First, a character string in the text data is searched (S203), and it is confirmed whether or not there is a word related to the extracted item in the first order (S204). If there is a related word, the range to be displayed is determined and displayed in the second display area (details will be described later). If there is no related word, the item is treated as not existing (the target item information column in the second display area is left blank), and the process continues with the related words for the extracted item in the next order (S205, S206). This is repeated until there are no extracted items in the next order, and at that point, the extraction of target item information is terminated, the extracted character string is confirmed on the screen on which it is displayed, and if correction is required, the specific character string is corrected. The correction is performed using the input function of the terminal device. If there is no instruction to correct the character string (S207), the structuring process is terminated to confirm the information to be aggregated.

[0046] On the other hand, if a correction instruction is given (S207), the specified character string is reflected in the text data (S208). When the corrected character string is reflected in the text data, the originally assigned ordinal number is updated, a new ordinal number is assigned to the entire character string, and processing is performed so that the ordinal numbers of the character strings displayed in the first and second display areas match. The corrected character string and ordinal number are displayed in both display areas with the updated content, and if further correction is required, the above processing is repeated. Then, when the correction instruction is no longer given, the structuring processing is terminated.

[0047] Fig. 7 shows only the process flow for extracting target item information when a word related to an extraction item is detected in the above structuring process. This process flow is for determining whether or not to extract a character string as target item information when a predetermined word (and related words) exists as an extraction item. The process starts from process A in Fig. 6, and after the process, the process returns to process B in Fig. 6 to continue the structuring process.

[0048] As shown in Fig. 7, if there is a related word in the extracted item, it is determined whether the word is related to "body temperature," "high blood pressure," "low blood pressure," "physical condition," etc. (S301-S305). If the word is "body temperature," the next numerical part (e.g., 36.6) to "degrees" is extracted (S306). Also, for items that are only shown in numbers, such as blood pressure or pulse rate, the next numerical part of the word is extracted (S307, S308). The character string extracted as above is displayed in a predetermined target item information column in the second display area.

[0049] Also, if the word is "physical condition", the part from the word to an auxiliary verb such as "desu" is extracted, or if punctuation is added as the speech recognition result (character string), the part from the word to the first period "." is extracted and displayed in the target item information column of the corresponding extracted item (S308). Furthermore, if the extracted item is related to the situation of the "today", referring to related words meaning "today", the part from related words specifying the time period (such as "today" or "present") to an auxiliary verb such as "desu" (or to the period ".") is extracted and displayed in the second display area (S309). If it is related to future plans, referring to related words meaning the future such as "tomorrow" or "days later", the part from related words specifying the time period (such as "tomorrow" or "two days later") to an auxiliary verb such as "desu" (or to the period ".") is extracted and displayed in the second display area (S310).

[0050] In addition, when there are multiple items such as physical condition, the current day, and future plans, multiple items may be extracted and displayed in the target item information column of the second display area. Unnecessary information can be deleted by the above-mentioned correction instruction. Details of this type of processing method will be described later in the modified example.

[0051] <Summary> As described above, the embodiment of the present invention can display the entire character string obtained by converting the input voice into text in the first display area, and can display the target item information, which is the information to be aggregated, in the second display area, and can easily create a character string to be recorded based on the voice input and input it to a predetermined storage destination. Also, since the target item information is information related to a predetermined extraction item, if a character string converted into text from the input voice is obtained, it will inevitably be reflected in the extraction item, which contributes to reducing the effort of inputting.

[0052] This embodiment also illustrates a method for inputting patient information such as nursing information into a storage means such as an electronic medical record, and by constructing an environment in which medical information can be input into an electronic medical record, it becomes possible to easily store medical information in the electronic medical record. Note that this is not limited to inputting information into an electronic medical record, and it can also be used as an input support for creating minutes of a meeting where only formal items are decided.

[0053] <Modification> Although the present embodiment is as described above, the present invention is not limited to these embodiments, and each element of the above embodiment may be changed or other elements may be added. For example, when correcting a mistaken speech during voice input, items related to the same extraction item may be input multiple times, and in that case, the text data may be in a state in which, for example, information on "body temperature" is repeated twice, as shown in FIG. 8. In addition, there may be multiple target item information for the same extraction item, and multiple pieces of information should be extracted for the same extraction item. Assuming such a case, as a structuring process when there are multiple identical extraction items, information extraction is performed several times, and strings with small ordinal numbers are extracted in order (in the order of input), and target item information columns 22a to 22h can be added as appropriate, and these can be set to be displayed in the second display area 12.

[0054] In addition, when the amount of information extracted in the second display area 12 is large, the second display area 12 may be constructed to be scrollable. In addition, since the extracted target item information may include unnecessary information (information due to a slip of the tongue) or information that should be excluded from the information to be consolidated (saved), a delete key may be added to the target item information columns displayed in order to facilitate the deletion instruction. Furthermore, the deletion of each target item information column may not be processed as a correction of text data, but may be simply processed as a deletion of the display in the second display area 12 (cancellation of a display instruction), so that the original text data is saved without being modified. This type of deletion is not limited to the case where multiple target item information exists in the same extracted item, and a delete key may be provided for all target item information to enable deletion of each item (cancellation of a display instruction).

[0055] The processing flow of the above-mentioned modified example is shown in FIG. 9 and FIG. 10. Note that FIG. 9 corresponds to FIG. 6 used in explaining the above-mentioned embodiment, and FIG. 10 corresponds to FIG. 7. To explain the modified example with reference to these figures, as shown in FIG. 9, in the modified example, during the structuring process of FIG. 6, it is determined whether the related word is singular or plural. That is, it is determined whether the related word extracted for one extraction item is the first word (S401), and if it is the first word, as in the case of the embodiment (FIG. 6), a predetermined target item information is displayed in the second display area. After this process is completed, it is determined whether all related words have been extracted (S402). If the related word is singular, all related words have been extracted, so the process moves to the process for the next extraction item, but if it is plural, the second related word is extracted, so the process for the second related word is executed (moves to "C" to "D" (FIG. 10)). The above process is repeated until all related words are extracted. If there is a related word in the process when there is a next extracted item (S205), the same process is carried out for all extracted items while changing the process according to a single or multiple related words in the same manner as above.

[0056] As shown in Fig. 10, when the second or subsequent related words are extracted, the process is basically the same as when the first related word is extracted (Fig. 7), and it is determined whether the word is related to "body temperature", "high blood pressure", "low blood pressure", "physical condition", etc. (S501 to S505). If the word is "body temperature", the next numerical part (e.g., 36.6) to "degrees" is extracted (S506), and if the word is expressed only by numbers such as blood pressure or pulse rate, the next numerical part of the word is extracted (S507, S508). If the word is related to "physical condition" or the state of the day, etc., the part from the word to an auxiliary verb such as "desu" or the first period "." is extracted (S508 to S510). When these are extracted, since the target item information for the first related word is already displayed in the second display area (target item information column), the next target item information column is added and then the extracted target item information is displayed (S506 to S510).

[0057] By the above-described processing, it is possible to extract multiple pieces of information (target item information) related to the extracted items, and by making it possible to delete the target item information column as a unit only when it is unnecessary, ultimately only the information to be aggregated (stored) is displayed in the second display area, and by sending the displayed information to the storage destination, assistance in inputting the appropriate information to be stored is completed.

[0058] In the above embodiment or modified example, the final aggregated information (aggregated information composed of extracted items and target item information) can be an extraction-type summary in that state, and can therefore be used as one form of summary. In addition, it is also possible to create a generative summary by, for example, processing the target item information (character strings) and codifying it into a literary style. Generative AI technology may be used to process this type of character string. In particular, when creating a summary to be used in meeting minutes, etc., it is expected that editing it into a generative summary will make it easier to verify at a later date.

[0059] In the above embodiment and modified examples, a smartphone is exemplified as the terminal device 1, and it is assumed that the terminal device 1 is equipped with an arithmetic processing unit, memory, storage, a microphone, a speaker, a communication interface, a touch screen, and the like. However, the terminal device is not limited to a smartphone, and any device equipped with similar functions can be used as the terminal device, and may be a general-purpose PC that can be connected to a network, a tablet PC, or a notebook PC in an environment where data can be transmitted and received. Although wireless data transmission and reception is convenient, wired data transmission and reception may also be used.

[0060] Furthermore, for items to be corrected, the target item information displayed in the second display area 12 will be the focus, but since important elements may also be included in the text data displayed in the first display area 11, corrections may be made mainly to the character strings displayed in the first display area 11. When both are saved, adjusting the character strings displayed in the first display area 11 will enable them to be clearly read when checked at a later date. [Explanation of symbols]

[0061] 1 Terminal equipment 2 Network 3. API (Web Speech API) 4 Structured Server 5 Electronic medical records (electronic medical records system) 10 Display screen 11 First display area 12 Second display area 21 Display column (for text data) 22a~22h Target item information column

Claims

1. An input support system comprising a terminal device capable of transmitting and receiving voice, a voice recognition means for recognizing input voice and converting it into text data, and a processing means for processing the text data recognized by the voice recognition means, the input support system supporting input to a predetermined storage means, the terminal device comprises: an input means for inputting voice information; a storage means for storing the voice data and the text data inputted by the input means; and a display means for displaying the text data; the display means displays a first display area for displaying the text data and a second display area for displaying information obtained by extracting only desired elements from the text data; The processing means displays the entire text data in the first display area, extracts item target information related to a predetermined extraction item from the text data, and displays the item target information in the second display area, and when a character string displayed in one of the first and second display areas of the terminal device is corrected, the processing means processes the correction of the same character string displayed in the other display area.

1. An input support system comprising:

2. The processing means includes: assigning an ordinal number to each character of the character string constituting the text data, and storing each character together with the ordinal number in the storage means; displaying, in the first display area, a character string in which each character is arranged in accordance with the order of the ordinal numbers stored in the storage means, and displaying, in the second display area, a character string in which each character constituting the item target information is arranged based on the order of the ordinal numbers assigned to the character; When a part of characters displayed in either the first or second display area is corrected, the ordinal number for the entire character string after the correction is updated and reassigned, and the information stored in the storage means is updated. The input support system according to claim 1 .

3. 3. The input support system according to claim 2, wherein the voice recognition means is provided in an external device capable of transmitting and receiving data from the terminal device, and the terminal device temporarily stores text data received from the external device in the storage means.

4. 4. The input support system according to claim 3, wherein said processing means, when detecting a plurality of pieces of item target information, sequentially displays all of the plurality of pieces of item target information in said second display area.

5. An input support system for medical records using the input support system according to any one of claims 1 to 4, The storage means stores examination information, treatment information and other medical management information for each patient, the terminal device is capable of accessing the medical information for each patient stored in the storage means; A medical record input support system characterized in that the processing means processes items indicating medical factors as the extracted items and information to be recorded as the medical information for each patient as the item target information.

6. the storage means is provided to be capable of storing the text data displayed in the first display area and the item target information displayed in the second display area, 6. The medical record input support system according to claim 5, wherein said processing means processes said text data and said item target information so as to transmit either one or both of them to said storage means while distinguishing between them.

Images