Patient result reporting method using artificial intelligence-based voice and apparatus for performing method

Abstract

The present invention relates to a voice-based patient result self-reporting method using artificial intelligence and an apparatus for performing the method. The voice-based patient result self-reporting method using artificial intelligence may comprise the steps in which: a patient analysis device receives basic patient analysis information; the patient analysis device generates basic report information on the basis of the basic patient analysis information; and a patient report generation device generates a patient report on the basis of the basic report information. The present invention is technology developed through the Seoul Industrial Development Institute 2021 Bio-Medical Technology Commercialization Support Project (BT210029) "Development of Alcohol Addiction Precision Digital Therapeutic Agent Technology".

Description

Method for reporting patient outcomes using artificial intelligence-based voice and device for performing the method

[0001] The present invention relates to a method for reporting patient outcomes using artificial intelligence-based voice and a device for performing such a method. More specifically, the present invention relates to a method for reporting patient outcomes using artificial intelligence-based voice for generating patient reports for various purposes and a device for performing such a method.

[0002] The development of various smart technologies has enabled the recording of data on individuals' daily activities, enabling them to manage their lives more efficiently. Among these, health-related data logging is gaining attention due to the growing interest in health. Many users are already generating and utilizing various health-related data, such as exercise, diet, and sleep, through their smartphones and wearable devices. This shift is a shift from the past, when health-related data was solely generated and managed by medical institutions, to users themselves, through their own devices like smartphones and wearables.

[0003] Health-related data logging is often accomplished through wearable devices. Wearable devices are devices worn or attached to the user's body. Thanks to advancements in technologies like the Internet of Things, wearable devices are increasingly being used to collect health-related data. Wearable devices can collect information about the user's physical changes and the environment surrounding them, and based on this data, they can provide necessary health advice.

[0004] A user's health-related data may include user biomarkers, and research is underway on methods to provide more adaptive medical prescriptions to users based on their health-related data.

[0005] This invention is a technology developed through the Seoul Business Agency's 2021 Bio-Medical Technology Commercialization Support Project (BT210029) "Development of Precision Digital Treatment Technology for Alcohol Addiction."

[0006] Related technology includes Korean registered patent No. 10-2425479.

[0007] The purpose of the present invention is to solve all of the above-described problems.

[0008] In addition, the present invention aims to determine a user's disease condition based on the user's voice data and generate a report on the user's condition.

[0009] In addition, the present invention aims to provide patient information for various targets (guardians, doctors, insurance companies) using only the user's voice by generating a patient report based on the user's voice data.

[0010] A representative configuration of the present invention to achieve the above purpose is as follows.

[0011] According to one embodiment of the present invention, a voice-based patient self-reporting method using artificial intelligence may include a step in which a patient analysis device receives patient analysis basic information, a step in which the patient analysis device generates report basic information based on the patient analysis basic information, and a step in which a patient report generation device generates a patient report based on the report basic information.

[0012] Meanwhile, the patient analysis basic information includes voice data, and the patient analysis device can input the voice data into a voice data analysis artificial intelligence engine to generate the report basic information.

[0013] In addition, the patient analysis device may further include a query generation artificial intelligence engine for obtaining necessary information from a user when information required for the voice data analysis artificial intelligence engine exists.

[0014] According to another embodiment of the present invention, a patient report system for voice-based patient self-result reporting using artificial intelligence may include a patient analysis device implemented to receive patient analysis basic information and generate report basic information based on the patient analysis basic information, and a patient report generation device implemented to generate a patient report based on the report basic information.

[0015] Meanwhile, the patient analysis basic information includes voice data, and the patient analysis device can input the voice data into a voice data analysis artificial intelligence engine to generate the report basic information.

[0016] In addition, the patient analysis device may further include a query generation artificial intelligence engine for obtaining necessary information from a user when information required for the voice data analysis artificial intelligence engine exists.

[0017] According to the present invention, the user's disease condition can be determined based on the user's voice data, and a report on the user's condition can be generated.

[0018] In addition, according to the present invention, patient information for various targets (guardians, doctors, insurance companies) can be provided using only the user's voice by generating a patient report based on the user's voice data.

[0019] Figure 1 is a conceptual diagram illustrating a patient report system according to an embodiment of the present invention.

[0020] Figure 2 is a conceptual diagram illustrating a patient analysis device according to an embodiment of the present invention.

[0021] Figure 3 is a conceptual diagram illustrating a voice data analysis artificial intelligence engine according to an embodiment of the present invention.

[0022] Figure 4 is a conceptual diagram illustrating a report generation method according to an embodiment of the present invention.

[0023] Figure 5 is a conceptual diagram illustrating a report generation method according to an embodiment of the present invention.

[0024] Figure 6 is a conceptual diagram illustrating a method for generating patient report information according to an embodiment of the present invention.

[0025] FIG. 7 is a conceptual diagram illustrating a method for generating a patient report based on a user's voice data according to an embodiment of the present invention.

[0026] The following detailed description of the present invention refers to the accompanying drawings, which illustrate specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention, while different from each other, are not necessarily mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be modified and implemented from one embodiment to another without departing from the spirit and scope of the present invention. Furthermore, it should be understood that the positions or arrangements of individual components within each embodiment may also be modified without departing from the spirit and scope of the present invention. Accordingly, the following detailed description is not to be taken in a limiting sense, and the scope of the present invention is to be construed to encompass the scope of the claims and all equivalents thereof. Like reference numerals in the drawings represent the same or similar elements throughout the several aspects.

[0027] Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the attached drawings so that a person having ordinary skill in the art to which the present invention pertains can easily practice the present invention.

[0028]

[0029] Hereinafter, for the convenience of explanation, the present invention is described assuming that the user is a patient, but the user may be interpreted to include not only a patient but also a non-patient who has not been diagnosed with a specific disease, and such embodiments may also be included in the scope of the present invention.

[0030]

[0031] Figure 1 is a conceptual diagram illustrating a patient report system according to an embodiment of the present invention.

[0032] In Fig. 1, a patient report system for generating a patient report based on artificial intelligence is disclosed.

[0033] Referring to FIG. 1, the patient report system may include a user device (100), a patient analysis device (120), a patient report generation device (140), and an information security device (160).

[0034] The user device (100) may be a device for transmitting patient analysis basic information, such as user voice data, which is the basis for patient analysis, to the patient analysis device.

[0035] The patient analysis device (120) may be implemented to generate report base information for generating patient reports based on patient analysis base information transmitted from the user. For example, the patient analysis device (120) may generate report base information for generating reports for various targets (e.g., doctors, guardians, insurance companies) based on the user's voice information.

[0036] A patient report generation device (140) may be implemented to generate a patient report. The patient report generation device (140) may generate a patient report based on report basic information generated by the patient analysis device. For example, the patient report generation device (140) may generate a first patient report for a doctor, a second patient report for a patient's guardian, and a third patient report for an insurance company.

[0037] An information security device (160) may be implemented to secure information generated from a user device (100), a patient analysis device (120), and a patient report generation device (140).

[0038]

[0039] Figure 2 is a conceptual diagram illustrating a patient analysis device according to an embodiment of the present invention.

[0040] In Fig. 2, a patient analysis device implemented to generate report basic information based on patient analysis basic information is disclosed.

[0041] Referring to FIG. 2, the patient analysis device can receive voice data as basic patient analysis information. The patient analysis device preprocesses the voice data, and the preprocessed voice data can be input into the voice data analysis artificial intelligence engine (200). The voice data analysis artificial intelligence engine (200) can generate report basic information based on the preprocessed voice data.

[0042] The voice data analysis AI engine (200) processes preprocessed voice data and generates report basic information. It can generate different report basic information for each report. For example, the voice data analysis AI engine (200) can generate first report basic information for a first patient report, second report basic information for a second patient report, and third report basic information for a third patient report.

[0043] In addition, the voice data analysis artificial intelligence engine (200) may be divided into a voice data analysis artificial intelligence engine (type 1) (210) for preferentially judging users without a record of previous medical history and a voice data analysis artificial intelligence engine (type 2) (220) for users with a record of previous medical history.

[0044] Separate previous medical history data is not input into the voice data analysis artificial intelligence engine (type 1) (210), and separate previous medical history data is additionally input into the voice data analysis artificial intelligence engine (type 2) (220), so that report basic information that additionally takes previous medical history data into consideration can be generated.

[0045] Additionally, according to an embodiment of the present invention, if information required for the voice data analysis artificial intelligence engine (200) exists, a query generation artificial intelligence engine (230) for obtaining the required information from a user may be additionally included in the patient analysis device.

[0046] The query generation artificial intelligence engine (230) can generate a query by considering 1) parts where it is difficult to fill in specific information in a report based on the user's voice data, 2) the reliability range for answers that are generated from a report using the user's voice data but have low reliability, etc.

[0047] That is, if information required for report generation for a specific patient report (Patient 1 Report, Patient 2 Report, Patient 3 Report) exists or information has been acquired but the reliability of the information is low, the query generation engine (230) can generate an inquiry for that section and transmit it to the user device. The user can generate a response to the inquiry as voice data and transmit it through the user device.

[0048] That is, the query generation engine (230) can generate a query and provide it to the patient by considering the type of report required, the information required for the report, and the reliability based on the patient's voice data.

[0049] Voice data can be input again into the voice data analysis artificial intelligence engine (200) and converted into report basic information required for report generation.

[0050]

[0051] Figure 3 is a conceptual diagram illustrating a voice data analysis artificial intelligence engine according to an embodiment of the present invention.

[0052] In Fig. 3, the operation of the voice data analysis artificial intelligence engine (type 1) and the voice data analysis artificial intelligence engine (type 2) is initiated.

[0053] Referring to FIG. 3, the specific operations of the voice data analysis artificial intelligence engine (type 1) (310) and the voice data analysis artificial intelligence engine (type 2) (320) are disclosed.

[0054] The voice data analysis artificial intelligence engine (type 1) (310) can perform a full screening for overall possible diseases based on user voice data, and generate report basic information for determining the possibility of various diseases based on the full screening.

[0055] The voice data analysis artificial intelligence engine (type 2) (320) can generate report basic information for creating a patient report on a disease that the user already has based on the user's voice data.

[0056] For users who have been previously diagnosed with a disease, report basic information can be generated through the voice data analysis artificial intelligence engine (type 1) (310) and the voice data analysis artificial intelligence engine (type 2) (320).

[0057] For users who have not been previously diagnosed with a disease, report basic information can be generated through a voice data analysis artificial intelligence engine (type 1) (310).

[0058] That is, it is possible to generate patient reports based on voice data, but it is also possible to generate patient reports based on existing diseases.

[0059] The voice data analysis artificial intelligence engine (Type 1) (310) can receive patient analysis basic information (Type 1) (315) when generating patient analysis information. The patient analysis basic information (Type 1) (315) can be generated by preprocessing sentences and words that can serve as sources for generating patient reports, and adding tags of all diseases highly related to the sentences and words to the sentences and words. At this time, the sentences and words can be grouped based on tags of each of multiple diseases. For example, sentences and words related to migraine can be preprocessed to include tags of other diseases that consider not only migraine but also the possibility of other diseases (e.g., arteriosclerosis).

[0060] For a report related to migraine, sentences and words related to the migraine tag are grouped into a migraine report group, and the migraine report group is input into a voice data analysis artificial intelligence engine (type 1) (310) and can be utilized to generate a report.

[0061] In this way, sentences and words related to each of a plurality of diseases are grouped based on tags for each of a plurality of diseases and created as a disease n report group, and the disease n report group is input into a voice data analysis artificial intelligence engine (type 1) (310), and the voice data analysis artificial intelligence engine (type 1) (310) can create report basic information for a target disease requiring a report among a plurality of diseases (disease 1 to disease n).

[0062] The voice data analysis artificial intelligence engine (type 2) (320) can receive patient analysis basic information (type 2) (325) when generating patient analysis information. The patient analysis basic information (type 2) (325) can include data with tags for sentences and words related to a target disease added to report the target disease that the user has. The voice data analysis artificial intelligence engine (type 2) (320) can generate report basic information for the target disease based on the patient analysis basic information (type 2) (325) including tags for the target disease.

[0063] For example, depending on the type of patient report to be generated, first report basic information, second report basic information, and third report basic information may be generated. The first report basic information, second report basic information, and third report basic information may be input into the patient report generation engine to generate the first patient report, second patient report, and third patient report. The first patient report may be for a doctor, the second patient report may be for a guardian, and the third patient report may be for an insurance company.

[0064]

[0065] Figure 4 is a conceptual diagram illustrating a report generation method according to an embodiment of the present invention.

[0066] Figure 4 discloses a method for determining patient analysis baseline information for report generation. In particular, a method for generating patient analysis baseline information (Type 2) is disclosed for cases where information on a user's existing disease exists.

[0067] Referring to Fig. 4, the user's voice data can be converted into text. The text-converted voice data can be segmented into multiple sentences. Considering the disease information the user already has, some of the multiple sentences can be extracted as patient analysis basic sentences (420) based on a disease vocabulary group (disease n) (400). The disease vocabulary group (disease n) (400) may be a group including words that are highly related to symptoms of a specific disease. That is, among the text-converted voice data, sentences that include words corresponding to the disease vocabulary group (disease n) (400) or sentences that are highly related to the disease vocabulary group (disease n) (400) can be extracted as patient analysis basic sentences (420).

[0068] After the patient analysis basic sentences (420) are extracted, the patient analysis basic sentences (420) can be rearranged through the relationship between the patient analysis basic sentences (420).

[0069] The patient analysis basic sentences (420) can be classified by symptom based on the symptom-specific sub-disease vocabulary groups (430) included in the disease vocabulary group (disease n). For example, disease n includes symptom 1, symptom 2, and symptom 3, and the disease vocabulary group (disease n) (400) can include a sub-disease vocabulary group (symptom 1) for symptom 1, a sub-disease vocabulary group (symptom 2) for symptom 2, and a sub-disease vocabulary group 3 (symptom 3) for symptom 3 as the symptom-specific sub-disease vocabulary groups (430). The patient analysis basic sentences (420) can be classified into symptom-specific patient analysis basic sentences (440) based on the sub-disease vocabulary groups (430). For example, the patient analysis basic sentences (440) can be classified into patient analysis basic sentences (symptom 1), patient analysis basic sentences (symptom 2), and patient analysis basic sentences (symptom 3). The classified patient analysis basic sentence (symptom n) can be determined as a patient analysis basic sentence group (symptom n) (440).

[0070] The patient analysis basic sentence group (symptom n) (440) can be rearranged considering temporal sequence. If a word that can determine temporal sequence exists within a sentence, the patient analysis basic sentence (symptom n) can be rearranged within the patient analysis basic sentence group (symptom n) (440) considering temporal sequence.

[0071] Patient analysis basic information (type 2) (450) can be generated based on a group of patient analysis basic sentences (symptom n) (440) and temporally rearranged patient analysis basic sentences (symptom n). Current user symptom data is extracted based on the patient analysis basic sentences (symptom n), and the user symptom data can be used to generate patient analysis basic information (type 2) (450). In addition, if a relationship exists between symptoms, patient analysis basic information (type 2) (450) can be generated by considering the relationship between patient analysis basic sentences (symptom n) corresponding to different symptoms.

[0072] The patient analysis basic information (Type 2) (450) generated in the above manner can be input into a voice data analysis artificial intelligence engine (Type 2) to generate report basic information. The report basic information may be information that classifies the patient analysis basic information (Type 2) (450) into detailed items that should be reported.

[0073]

[0074] Figure 5 is a conceptual diagram illustrating a report generation method according to an embodiment of the present invention.

[0075] Figure 5 discloses a method for determining patient analysis baseline information for report generation. In particular, a method for generating patient analysis baseline information (Type 1) for cases where information on a user's existing disease does not exist is disclosed.

[0076] Referring to FIG. 5, in the case of the patient analysis basic information (type 2) described in FIG. 4, the patient analysis basic information (type 2) can be generated without a disease determination procedure because the report is generated based on the patient's disease that has already been determined.

[0077] In order to generate patient analysis basic information (type 1) (580), a disease determination procedure that primarily determines the user's disease can be performed.

[0078] (1) Disease assessment procedure

[0079] The user's voice data can be converted to text. The converted voice data can be segmented into multiple sentences. In the case of patient analysis basic information (Type 1) (580), since there is no user disease data for the user's previously confirmed disease, sentences with a high correlation to the symptom vocabulary group (500) among the multiple sentences can be classified as patient analysis basic sentences (520).

[0080] Considering the symptoms, some of the sentences among the plurality of sentences can be classified as patient analysis basic sentences (520) based on the symptom vocabulary group set (500). The symptom vocabulary group set (500) can include a plurality of symptom vocabulary groups (530) that group vocabulary related to symptoms. The symptom vocabulary group (530) is a group of vocabulary with high relevance to each symptom and can be defined for various symptoms. If the symptom is a headache, vocabulary such as head ache, migraine, and bone tingling can be included in the corresponding symptom vocabulary group (530). A single patient analysis basic sentence (520) can also correspond to different symptom vocabulary groups (530).

[0081] In the present invention, a symptom vocabulary group (530) corresponding to a patient analysis basic sentence is determined, and a prediction of the possibility of a disease can be performed based on the determined symptom vocabulary group (530). For example, if a patient analysis basic sentence (520) corresponds to symptom vocabulary group a, symptom vocabulary group b, and symptom vocabulary group c, a disease that can have all of symptom vocabulary group a, symptom vocabulary group b, and symptom vocabulary group c can be determined as a candidate predicted disease (540).

[0082] After determining the candidate predicted disease (540), the final predicted disease (560) can be determined based on the patient analysis basic sentences (520) corresponding to the symptom vocabulary groups (530). The final predicted disease (560) can be determined by considering the relationship between the patient analysis basic sentences, the relevance of words included in the patient analysis basic sentences to a specific disease, and the severity of specific symptoms included in the patient analysis basic sentences.

[0083] (2) Creation of patient analysis basic information (Type 1) (580)

[0084] When the final predicted disease is determined, the patient analysis basic sentence group (symptom n) (570) for the final predicted disease can be determined in the same manner as the procedure described above in FIG. 4, and patient analysis basic information (type 1) (580) can be generated based on the patient analysis basic sentence group (symptom n) (570).

[0085] If a candidate predicted disease (540) does not exist, patient analysis basic information (type 1) (580) can be generated based only on patient analysis basic sentences (520) corresponding to specific symptoms.

[0086] The patient analysis basic information (Type 1) (580) generated in the above manner can be input into a voice data analysis artificial intelligence engine (Type 1) to generate report basic information. The report basic information may be information that classifies the patient analysis basic information (Type 1) (580) into detailed items that should be reported.

[0087] According to an embodiment of the present invention, when a conflict occurs between patient analysis baselines, the reliability of the patient analysis baselines can be determined to address the conflict. When a conflict occurs, a report can be generated based on patient analysis baselines with relatively high reliability. The reliability of the patient analysis baselines can be determined by considering factors such as the degree of matching with previous patient analysis baselines and the degree of overlap between the patient analysis baselines.

[0088] Additionally, depending on the report target (e.g., doctor, guardian, insurance company), the reliability of the patient analysis basic information may be considered, and patient analysis basic information below a certain reliability level may be set not to be included in a specific report.

[0089]

[0090] Figure 6 is a conceptual diagram illustrating a method for generating patient report information according to an embodiment of the present invention.

[0091] In Fig. 6, a method for generating patient report information based on time-series information is disclosed.

[0092] Referring to FIG. 6, the patient report generation engine can generate a patient report by considering the relationship between existing patient analysis basic information.

[0093] The patient analysis basic information for generating a patient report can be classified into patient analysis basic information (time series) (600), which is time-series information, and patient analysis basic information (non-time series) (610), which is non-time-series information. The patient analysis basic information (time series) (600) may be information for generating a patient report by additionally considering previously received patient analysis basic information. The patient analysis basic information (non-time series) (610) may be information for generating a patient report by considering only the current patient analysis basic information.

[0094] In the case of patient analysis basic information (time series) (600), it is determined whether previous patient analysis basic information exists, and if so, the previous patient analysis basic information can be reflected when generating the current report. The time series unit of the patient analysis basic information (time series) (600) can be adaptively determined considering the accuracy of the report results and disease characteristics.

[0095] The minimum data unit for generating a report can be determined as a reference patient analysis time series unit (620) of patient analysis basic information (time series) (600). The reference patient analysis time series unit (620) can be defined in consideration of disease characteristics. For a specific disease, the time required for accumulation of patient analysis basic information for diagnosis or reporting can be defined, and the reference patient analysis time series unit (620) can be defined based on this. In addition, the reference patient analysis time series unit (620) can be defined to have a report result accuracy greater than a threshold. The report result accuracy (630) can be determined in consideration of whether the items generated as a report match the actual user's status.

[0096] The reference patient analysis time series unit (620) may be adjusted in consideration of the report result accuracy (630). In an embodiment of the present invention, the characteristics of the reference patient analysis time series unit (620) may be changed in consideration of the report result accuracy (630). For example, even when the amount of patient analysis basic information is relatively reduced or the time series unit of the patient analysis basic information is reduced due to an increase in the performance of the artificial intelligence engine and an increase in the report result accuracy (630), if it is possible to generate a report result with an accuracy higher than a threshold value, the reference patient analysis time series unit (620) may be reduced. That is, the characteristics of the reference patient analysis time series unit (620) may be newly set based on feedback on the report result.

[0097]

[0098] FIG. 7 is a conceptual diagram illustrating a method for generating a patient report based on a user's voice data according to an embodiment of the present invention.

[0099] In FIG. 7, a method for generating a patient report based on contextual characteristics and extra-contextual characteristics from voice data input by a user is disclosed.

[0100] Referring to FIG. 7, contextual and extra-contextual characteristics are classified from voice data input by a user, and based on the contextual data (700) and extra-contextual data (705), a disease (e.g., dementia, migraine, etc.) the user has can be predicted and the disease can be monitored. A patient report can be generated based on the disease prediction and disease monitoring results.

[0101] The preprocessing unit can extract contextual data (700) and non-contextual data (705) based on speech data. Contextual data (700) may include information on characteristics occurring in the context, such as sentence completeness, word structure, and vocabulary. Non-contextual data (705) may include information on characteristics outside the context (e.g., tone, pitch, degree of stuttering, etc.).

[0102] The voice data analysis AI engine (750, 760) can receive contextual data (700) and extra-contextual data (705) and generate user symptom data. The user symptom data may include disease prediction data and disease monitoring data. For example, a prediction of the user's likelihood of experiencing migraines may be performed based on the voice data analysis AI engine (750, 760), and migraine prediction data and migraine monitoring data may be provided as user symptom data.

[0103] Context data (700) may include multiple sub-context data, and non-context data (705) may include non-sub-context data.

[0104] In the present invention, there may be two reference sub-context data (first reference sub-context data (710), second reference sub-context data (730)) to extract the user's sub-context data.

[0105] The first reference sub-context data (710) may be sub-context data that is commonly shared by a critical percentage (e.g., 80%) or more of users with a specific disease. For example, the first reference sub-context data (710) may be sub-context data that is commonly shared by users with dementia or migraines. The first reference sub-context data (710) may adaptively change based on the accumulation of the user's voice data. The first reference sub-context data may also be set as a separate default value, taking into account the user's age, education level, gender, etc.

[0106] The second reference sub-context data (720) may be sub-context data excluding the first reference sub-context data (710). The second reference sub-context data (720) may be sub-context data observed in a specific user or a specific user group with a specific disease.

[0107] The user's voice data may be analyzed based on the first reference sub-context data (710) and the second reference sub-context data (720), and user symptom data may be generated based on the analysis results.

[0108] Similarly, in order to extract the user's sub-context data in the present invention, there may be two reference sub-context data (first reference sub-context data (730), second reference sub-context data (740)).

[0109] The first reference sub-context data (730) may be sub-context data that is commonly shared by a critical percentage (e.g., 80%) or more of users with a specific disease. For example, the first reference sub-context data (730) may be sub-context data that is commonly shared by users with dementia or migraines. The first reference sub-context data (730) may adaptively change according to the accumulation of the user's voice data. The first reference sub-context data (730) may also be set as a separate default value, taking into account the user's age, education level, gender, etc.

[0110] The second reference sub-context out-of-context data (740) may be sub-context out-of-context data excluding the first reference sub-context out-of-context data (730). The second reference sub-context out-of-context data (740) may be sub-context out-of-context data observed in a specific user or a specific user group with a specific disease.

[0111] The user's voice data may be analyzed based on the first reference sub-context out-of-context data (730) and the second reference sub-context out-of-context data (740), and user symptom data may be generated based on the analysis results.

[0112] According to an embodiment of the present invention, the voice data analysis artificial intelligence engine may be separately implemented as a voice data analysis artificial intelligence engine (first reference) (750) for analyzing first reference sub-context data (710) and first reference non-sub-context data (730), and a voice data analysis artificial intelligence engine (second reference) (760) for analyzing second reference sub-context data (720) and second reference non-sub-context data (740).

[0113] According to an embodiment of the present invention, user feature vectors (e.g., vocabulary vectors, sound feature vectors, etc.) of all users with a specific disease can be positioned on an embedding plane to determine first reference sub-context data (710) and first reference non-sub-context data (730). The user feature vectors can be vectors extracted based on the user's voice data.

[0114] As the user characteristic vector of user data (or user voice data) having a disease accumulates, the user characteristic vector commonly shared by users of a threshold criterion on the embedding plane changes, and accordingly, the vector corresponding to the first reference sub-context data (710) and the vector corresponding to the first reference sub-context extraneous data (730) may change. As described above, when the first reference sub-context data (710) and the first reference sub-context extraneous data (730) change according to the accumulation of the user characteristic vectors of all users having a specific disease, the learning data pool of the voice data analysis artificial intelligence engine (first reference) (750) and the voice data analysis artificial intelligence engine (second reference) (760) may change and be updated.

[0115]

[0116] The embodiments of the present invention described above may be implemented in the form of program commands that can be executed through various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, etc., either singly or in combination. The program commands recorded on the computer-readable recording medium may be specially designed and configured for the present invention or may be known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specifically configured to store and execute program commands, such as ROMs, RAMs, and flash memories. Examples of program commands include not only machine language codes generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. Hardware devices may be changed into one or more software modules to perform processing according to the present invention, and vice versa.

[0117] Although the present invention has been described above with specific details such as specific components and limited examples and drawings, these are provided only to help a more general understanding of the present invention, and the present invention is not limited to the above examples, and those with ordinary knowledge in the technical field to which the present invention pertains can make various modifications and changes based on this description.

[0118] Therefore, the idea of ​​the present invention should not be limited to the embodiments described above, and not only the scope of the patent claims described below but also all scopes equivalent to or equivalently modified from the scope of the patent claims are considered to fall within the scope of the idea of ​​the present invention.

Claims

1. A voice-based patient self-reporting method using artificial intelligence, A step in which a patient analysis device receives patient analysis basic information; A step in which the patient analysis device generates report basic information based on the patient analysis basic information; and A method characterized in that the patient report generation device includes a step of generating a patient report based on the report basic information.

2. In paragraph 1, The above patient analysis basic information includes voice data, A method characterized in that the patient analysis device inputs the voice data into a voice data analysis artificial intelligence engine to generate the report basic information.

3. In paragraph 2, A method characterized in that the patient analysis device further includes a query generation artificial intelligence engine for obtaining necessary information from a user when information required for the voice data analysis artificial intelligence engine exists.

4. A patient report system for voice-based patient self-reporting using artificial intelligence, A patient analysis device implemented to receive patient analysis basic information and generate report basic information based on the patient analysis basic information; and A patient report system characterized by including a patient report generation device implemented to generate a patient report based on the above report basic information.

5. In paragraph 4, The above patient analysis basic information includes voice data, A patient report system characterized in that the patient analysis device inputs the voice data into a voice data analysis artificial intelligence engine to generate the report basic information.

6. In paragraph 5, A patient report system characterized in that the patient analysis device further includes a query generation artificial intelligence engine for obtaining necessary information from a user when information required for the voice data analysis artificial intelligence engine exists.

Images