Method for managing biological information of user by hospital guide system based on artificial intelligence

Abstract

The invention discloses a method for managing biological information of a user by a hospital guide system based on artificial intelligence, which comprises the following steps: the user interacts with intelligent hospital guide equipment, and the intelligent hospital guide equipment binds collected audio information and facial image information with a user account of the user and stores the information in a corresponding user database; when the user uses the intelligent hospital guide equipment again and performs interaction, the intelligent hospital guide equipment calls data in the user database of the registered user from the server after comparing that the facial image information of the user is consistent with the facial image information of the user database, and the intelligent hospital guide equipment performs normal interaction with the registered user; collecting the current audio information of the registered user in real time; and when the deviation value between the current audio information of the registered user and the audio information in the user database is large, the intelligent hospital guide equipment actively sends out reminding information, so that a reminding function can be automatically performed, and the medical experience and safety of the patient are further improved.

Description

Technical Field

[0001] The present invention specifically relates to a method for managing user biometric information by a medical guidance system based on artificial intelligence. Background Art

[0002] With the growing demand for medical services and the rapid advancement of technology, hospitals are facing the important challenge of how to improve service quality and optimize patients' medical experience. The traditional medical process usually involves multiple links, including registration, waiting for the number to be called, consultation, examination and medication collection. This series of processes is not only time-consuming, but also often requires patients to run back and forth between different departments of the hospital, which brings great inconvenience to patients, especially elderly patients and those with limited mobility.

[0003] In order to solve these problems, many large hospitals have introduced intelligent electronic diagnosis and treatment systems and guidance equipment to improve patients' medical experience. For example, the "Hospital Diagnosis and Treatment Management System and Method" with patent number CN201910745328.9 simplifies the registration and treatment process by setting up self-service terminals, thereby improving efficiency. In addition, the "A Hospital Intelligent Medical Guidance Device and Method" with patent number CN201810093728.1 also provides an intelligent guidance device that can guide patients to quickly find the required departments and services.

[0004] However, although these intelligent systems and devices have achieved remarkable results in process optimization, they generally lack the monitoring and feedback functions of users' biological information, especially the monitoring of sound information. For example, when the patient's health status changes, the existing guidance equipment cannot detect these changes in time and provide corresponding suggestions or warnings. This is especially important for elderly patients and patients with chronic diseases, because they may experience drastic changes in their health status in a short period of time, and these changes can often be reflected by changes in sound. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method for managing user biometric information in a medical guidance system based on artificial intelligence, which can automatically provide reminders and further improve the patient's medical experience and safety.

[0006] The present invention is achieved through the following technical solutions:

[0007] The present invention discloses a method for managing user biometric information of a medical guidance system based on artificial intelligence, comprising a server, and an intelligent medical guidance device connected to the server in communication, wherein the intelligent medical guidance device is provided with an intelligent answering system for conducting voice dialogue with the user, and a display module for displaying AI virtual characters and text and picture information, and the intelligent medical guidance device is also provided with a camera device for collecting user image information, a sound receiving device for collecting user voice information, and a sound generating device for sound output;

[0008] The method for managing user biometric information by the artificial intelligence-based medical guidance system comprises the following steps:

[0009] S1. The user interacts with the intelligent medical guidance device. The user can register and establish a user account in the intelligent answering system. The system establishes a corresponding user database for the user account and uploads it to the server. The intelligent medical guidance device collects the user's audio information through the sound receiving device and the user's facial image information through the camera device, and uses the collected facial image information as authentication information to identify the user. The user who meets the authentication information is marked as a registered user, and the collected audio information and facial image information are bound to the user account of the user and stored in the corresponding user database;

[0010] S2. When the user uses the smart medical guidance device again and interacts with it, the smart medical guidance device compares the facial image information of the user with the facial image information in the user database, and then calls the data in the user database of the registered user from the server. The smart medical guidance device interacts with the registered user normally, and collects the current audio information of the registered user in real time.

[0011] S3. In step S2, the intelligent medical guidance device synchronously compares the current audio information of the registered user with the audio information in the user database. When the deviation value between the current audio information of the registered user and the audio information in the user database is large, the intelligent medical guidance device actively sends a reminder message.

[0012] The method for managing user biometric information of the medical guidance system based on artificial intelligence as described above is characterized by:

[0013] In step S1, the intelligent medical guidance device extracts the initial pitch value fo, the initial loudness value L0 and the initial formant characteristic value F0 in the audio information corresponding to the user account;

[0014] In step S3, after the intelligent medical guidance device collects the user's audio information, it extracts the current pitch value f, the current loudness value Ln and the current formant characteristic value F in the audio information;

[0015] In step S3, the pitch difference fx between the current pitch value f and the initial pitch value fo, the loudness difference Lx between the current loudness value Ln and the initial loudness value L0, and the peak feature difference Fx between the current resonance peak eigenvalue F and the initial resonance peak eigenvalue F0 are calculated respectively, and it is determined that the deviation between the current audio information of the registered user and the audio information in the user database is large when the following conditions are met: the pitch difference fx is greater than or less than the preset range of the pitch difference, and the loudness difference Lx is greater than or less than the preset range of the loudness difference, and the peak feature difference Fx is greater than or less than the preset range of the peak feature difference.

[0016] The method for managing user biometric information of the medical guidance system based on artificial intelligence as described above, the steps of the intelligent medical guidance device identifying and collecting user audio information include:

[0017] A. When a user approaches the intelligent medical guidance device and is recognized by the camera device, the intelligent medical guidance device recognizes and calculates the real-time distance between the user and the device. When the real-time distance is less than or equal to the first preset distance, the intelligent medical guidance device collects the user's facial image information;

[0018] B. The intelligent medical guidance device identifies and determines whether the user is wearing a mask. If the user is wearing a mask, the process proceeds to step C; if the user is not wearing a mask, the process proceeds to step D;

[0019] C. When the user wears a mask, the intelligent medical guidance device calculates the real-time distance between the user and the device. When the real-time distance is less than or equal to the second preset distance, the sound receiving device starts working and collects the user's voice information;

[0020] D. When the user is not wearing a mask, the intelligent medical guidance equipment calculates the user's lip movements through real-time facial image information to determine whether the user is speaking. When the user starts speaking, the sound receiving device starts working and collects the user's voice information.

[0021] The method for managing user biometric information of the medical guidance system based on artificial intelligence as described above, the step of extracting the pitch in the audio information includes:

[0022] A. performing frame processing on the audio information to obtain a plurality of framed audios;

[0023] B. Performing Fourier transform on each frame of the divided audio in step A to obtain frequency domain data, obtaining the peak point frequency corresponding to the peak point in the frequency domain data, and determining the average value of the peak point frequency as the fundamental frequency;

[0024] C. Based on the fundamental frequency in the framed audio, the pitch of the framed audio is calculated;

[0025] D. Generate a pitch value of the audio information based on the pitch of the framed audio.

[0026] The method for managing user biometric information of the medical guidance system based on artificial intelligence as described above, the step of extracting loudness from audio information includes:

[0027] A. The radio receiver captures the sound, converts it into an electrical signal and amplifies the electrical signal;

[0028] B. The intelligent diagnosis guidance device calculates the effective value of the electrical signal to obtain the average sound pressure P;

[0029] C. According to the formula SPL=log10(P / P0), where P0 is the reference sound pressure, for the sound pressure in the air, P0 is 20 microPascals (20μPa), and the calculated SPL is the loudness value.

[0030] The method for managing user biometric information of the medical guidance system based on artificial intelligence as described above, the step of extracting the resonance peak feature value in the audio information includes:

[0031] A. Acquire an audio signal, perform frame processing on the audio signal, and obtain N voice frames;

[0032] B. Perform fast Fourier transform on each speech frame to obtain the spectrum of each speech frame, obtain the formant data of each speech frame based on the spectrum, and take the first three formant data of each speech frame;

[0033] C. Calculate the statistical average of the three vibration peak data, which is the resonance peak characteristic value.

[0034] In the above-mentioned method for managing user biometric information of the medical guidance system based on artificial intelligence, the steps of the intelligent medical guidance device determining whether the user is speaking include:

[0035] D1. The camera device collects the user's facial image information, converts the facial image information into an image sequence, extracts the face area of ​​each frame in the image sequence, identifies the face part in the face image based on the convolutional neural network (CNN) architecture, and then obtains the face key points in the face frame area image, from which the upper lip feature point O and the lower lip feature point Q are extracted;

[0036] D2. Calculate the distance between the upper lip feature point and the lower lip feature point in each frame to obtain a distance change peak value. When the distance change peak value is greater than a preset peak value, it is determined that the user is speaking.

[0037] The method for managing user biometric information of the medical guidance system based on artificial intelligence as described above includes the following sub-steps in step D1:

[0038] D11, obtaining facial key points in the face frame area image, including the upper midpoint O1 of the upper lip, the lower midpoint O2 of the upper lip, the upper midpoint Q1 of the lower lip, and the lower midpoint Q2 of the lower lip;

[0039] D12. Extract the midpoint between the upper midpoint O1 of the upper lip and the lower midpoint O2 of the upper lip as the upper lip feature point O, and extract the midpoint between the upper midpoint Q1 of the lower lip and the lower midpoint Q2 of the lower lip as the lower lip feature point Q.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] The present invention provides a method for managing user biometric information in a medical guidance system based on artificial intelligence. The system can not only answer users' questions, but also provide personalized health advice by collecting and analyzing users' biometric information (such as facial images and voice information). By introducing a biometric information monitoring function based on artificial intelligence and a function of timely reminding when abnormalities occur in the user's biometric information, the patient's medical experience and safety are further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the principle of the server and several intelligent medical guidance devices of the present invention;

[0043] Figure 2 It is a flow chart of a method for managing user biometric information by an artificial intelligence-based medical guidance system of the present invention;

[0044] Figure 3 It is a schematic diagram of key points of a face in a face frame area image of the present invention. DETAILED DESCRIPTION

[0045] The invention will be further described below in conjunction with the accompanying drawings:

[0046] The directions described in the specification of the invention, such as "up", "down", "left", "right", "front", "back", etc., are based on the directions of the drawings and are intended to facilitate the description of the relationship between the various components. They do not indicate the unique or absolute positional relationship between the various components. They are only one of the implementation methods of the invention and are not a limitation on its implementation method.

[0047] A method for managing user biometric information in a medical guidance system based on artificial intelligence, such as Figure 1As shown, the system includes a server 1 and multiple intelligent medical guidance devices 2, and the intelligent medical guidance devices 2 and the server 1 are interconnected by wired or wireless means; the method for managing user biometric information based on an artificial intelligence medical guidance system is characterized in that it includes a server 1 and an intelligent medical guidance device 2 that is communicatively connected to the server 1, and the intelligent medical guidance device 2 is provided with an intelligent response system for conducting voice conversations with the user, and a display module 21 for displaying AI virtual characters and text and picture information. The intelligent medical guidance device 2 is also provided with a camera 22 for collecting user image information, a sound receiving device 23 for collecting user voice information, and a sound generating device 24 for sound output; wherein the intelligent response system is an intelligent response model constructed based on a recurrent neural network, and the intelligent response system can also refer to the patent number CN202410722371.4, and the name of a technical solution of a medical service platform guidance analysis method and system based on intelligent response.

[0048] The method for managing user biometric information by the artificial intelligence-based medical guidance system comprises the following steps:

[0049] S1. The user interacts with the intelligent medical guidance device 2. The user can register and establish a user account in the intelligent answering system. The system establishes a corresponding user database for the user account and uploads it to the server 1. The intelligent medical guidance device 2 collects the user's audio information through the sound receiving device 23 and the user's facial image information through the camera device 22, and uses the collected facial image information as authentication information to identify the user, and marks the user who meets the authentication information as a registered user. The collected audio information and facial image information are bound to the user account of the user and stored in the corresponding user database. For the use scenario of the intelligent medical guidance device 2, if there are multiple human targets, facial recognition is performed on the crowd image and facial image information of multiple individuals is extracted. The system automatically determines which human target is the user. For example, the human target closest to the user is determined as the user. For the method of obtaining facial image information of human targets, reference can be made to the patent scheme with patent number CN201910483312.5 and the name of a method and system for intelligent image optimization based on group image selection.

[0050] S2, when the user uses the smart medical guidance device 2 again and interacts with it, the smart medical guidance device 2 compares the facial image information of the user with the facial image information in the user database, and then calls the data in the user database of the registered user from the server 1, and the smart medical guidance device 2 interacts normally with the registered user, and collects the current audio information of the registered user in real time;

[0051] S3. In step S2, the intelligent medical guidance device 2 synchronously compares the current audio information of the registered user with the audio information in the user database. When the deviation value between the current audio information of the registered user and the audio information in the user database is large, the intelligent medical guidance device 2 actively sends out a reminder message. The reminder message includes a reminder or suggestion to the patient, for example, when the sound information occurs, it is suggested that "the sound information has abnormal changes, please pay attention to observation or seek medical treatment in time".

[0052] In this embodiment, the method for determining whether the deviation between the current audio information of the registered user and the audio information in the user database is large is as follows:

[0053] In step S1, the intelligent medical guidance device 2 extracts the initial pitch value fo, the initial loudness value L0 and the initial formant characteristic value F0 in the audio information corresponding to the user account;

[0054] In step S3, after the intelligent medical guidance device 2 collects the audio information of the user, it extracts the current pitch value f, the current loudness value Ln and the current formant characteristic value F in the audio information;

[0055] In step S3, the pitch difference fx between the current pitch value f and the initial pitch value fo, the loudness difference Lx between the current loudness value Ln and the initial loudness value L0, and the peak feature difference Fx between the current formant feature value F and the initial formant feature value F0 are calculated respectively. When the following conditions are met, it is determined that the deviation value between the current audio information of the registered user and the audio information in the user database is large: the pitch difference fx is greater than or less than the preset range of the pitch difference, and the loudness difference Lx is greater than or less than the preset range of the loudness difference, and the peak feature difference Fx is greater than or less than the preset range of the peak feature difference. Among them, the preset range of the pitch difference, the preset range of the loudness difference, and the preset range of the peak feature difference are all numerical ranges preset in the system of the intelligent medical guidance device 2. That is, the difference between the currently collected value and the recorded initial value exceeds the preset range. Among them, in this embodiment, three loudness, pitch and formant feature values ​​are extracted. Of course, only one or two of them can be extracted as parameters, and it can even be designed to extract 4 or more parameters, and it is not limited to three parameters.

[0056] Specifically, the step of extracting the pitch in the audio information includes:

[0057] A. performing frame processing on the audio information to obtain a plurality of framed audios;

[0058] B. Performing Fourier transform on each frame of the divided audio in step A to obtain frequency domain data, obtaining the peak point frequency corresponding to the peak point in the frequency domain data, and determining the average value of the peak point frequency as the fundamental frequency;

[0059] C. Based on the fundamental frequency in the framed audio, the pitch of the framed audio is calculated;

[0060] D. Generate a pitch value of the audio information based on the pitch of the framed audio.

[0061] Among them, taking the nth frame of framed audio as an example, the nth frame of framed audio in at least two frames of framed audio is Fourier transformed to obtain frequency domain data (spectrum), where n is a positive integer; the frequency domain data within the frequency range of the fundamental frequency of human voice in the frequency domain data is intercepted to obtain human voice frequency domain data, where the fundamental frequency of human voice can be a preset value, for example, the frequency range of the fundamental frequency of human voice is 50Hz to 500Hz; the peak point frequency corresponding to the peak point in the human voice frequency domain data is obtained; and the average value of the peak point frequencies is determined as the fundamental frequency of the nth frame of framed audio.

[0062] In addition, the preset range of the pitch difference is a system preset value, for example, the preset range of the pitch difference is 40 Hz to 100 Hz.

[0063] The steps of extracting loudness from audio information include:

[0064] A. The sound receiving device 23 captures the sound, converts it into an electrical signal and amplifies the electrical signal;

[0065] B. The intelligent diagnosis guidance device 2 calculates the effective value of the electrical signal to obtain the average sound pressure P;

[0066] C. According to the formula SPL=log10(P / P0), where P0 is the reference sound pressure, for the sound pressure in the air, P0 is 20 microPascals (20μPa), and the calculated SPL is the loudness value.

[0067] The preset range of the loudness difference is a system preset value, for example, the preset range of the loudness difference is 0 to 20 dB.

[0068] The steps of extracting the formant feature value in the audio information include:

[0069] A. Acquire an audio signal, perform frame processing on the audio signal, and obtain N voice frames;

[0070] B. Perform fast Fourier transform on each speech frame to obtain the spectrum of each speech frame, obtain the formant data of each speech frame based on the spectrum, and take the first three formant data of each speech frame;

[0071] C. Calculate the statistical average of the three vibration peak data, which is the resonance peak characteristic value.

[0072] Among them, sound information is collected and converted into a digital signal, and the digital signal is pre-emphasized to obtain an audio signal; the audio signal is framed to obtain N frames, the length of each frame is L, 20ms≤L≤50ms, the length of the overlapping area between two adjacent frames is M, M=(1-λ)*L, 0<λ≤0.5; each frame is windowed to obtain N speech frames. The sampling frequency of the sound information is 44.1kHz, the quantization bit number is 16 bits, the length of each frame is L=23.2ms, that is, 1024 sample points, the length of the overlapping area between two adjacent frames is M=13.2ms, that is, 583 sample points, and the specific framing process is: generate a sampling segment with a length of 23.3ms, the sample points falling within the sampling segment are a frame, the initial starting point of the sampling segment is the starting point of the audio signal, and the first frame is obtained. After that, the sampling segment moves 10ms (i.e. frame shift), i.e. 441 sample points, to obtain the second frame, the overlapping area between the first frame and the second frame is 1024-441=583, and the sampling segment is continued to be moved, each time by 10ms. In this way, 100 frames of sound information can be obtained every second; the length L of each frame is set to 1024 sample points, so that the computer is easier to calculate when performing fast Fourier transform later. There is an overlapping area between adjacent frames, which can enable a smooth transition between frames. In this embodiment, the frame shift is 10ms. In other implementations, the size of the frame shift can also be changed as needed to adjust the length M of the overlapping area between two adjacent frames; each frame is windowed using a Hamming window, specifically, each frame is multiplied by a window function of the Hamming window; each speech frame is fast Fourier transformed to obtain the spectrum of each speech frame, and the resonance peak data of each speech frame is obtained based on the spectrum, thereby obtaining the first 3 resonance peak data F1, F2, and F3 of each speech frame.

[0073] The preset range of the peak characteristic difference is the system preset value, for example, the preset range of the peak characteristic difference is 0-2500Hz

[0074] As a more preferred implementation mode of the present invention, the steps of the intelligent medical guidance device 2 identifying and collecting the user's audio information include:

[0075] A. When the user approaches the intelligent medical guidance device 2 and is recognized by the camera device 22, the intelligent medical guidance device 2 identifies and calculates the real-time distance between the user and the device. When the real-time distance is less than or equal to the first preset distance, the intelligent medical guidance device 2 collects the user's facial image information;

[0076] B. The intelligent medical guidance device 2 identifies and determines whether the user is wearing a mask. If the user is wearing a mask, the process proceeds to step C; if the user is not wearing a mask, the process proceeds to step D;

[0077] C. When the user wears a mask, the intelligent medical guidance device 2 calculates the real-time distance between the user and the device. When the real-time distance is less than or equal to the second preset distance, the sound receiving device 23 starts working and collects the user's voice information;

[0078] D. When the user is not wearing a mask, the intelligent medical guidance device 2 calculates the user's lip movements through real-time facial image information to determine whether the user is speaking. When the user starts speaking, the sound receiving device 23 starts working and collects the user's voice information.

[0079] Among them, the first preset distance is 2-5 meters, and the second preset distance is 0.1-1.9 meters.

[0080] Among them, regarding how the camera device 22 identifies and calculates the target distance: the camera device 22 uses a binocular camera, and derives the formula = f·b / d·ps based on the parallax principle to calculate the distance D1, where f is the focal length of the binocular camera, b is the baseline distance between the two cameras of the binocular camera, d is the parallax, and ps is the size of a single pixel in the image sensor.

[0081] Among them, regarding how to identify and determine whether the user is wearing a mask: you can refer to the patent scheme with patent number CN202010937955.5, named An adaptive portrait temperature matching and mask recognition method and device.

[0082] The steps for the intelligent medical guidance device 2 to determine whether the user is speaking include:

[0083] D1, the camera device 22 collects the user's facial image information, converts the facial image information into an image sequence, extracts the face area of ​​each frame in the image sequence, identifies the face part in the face image based on the convolutional neural network CNN architecture, and then obtains the face key points in the face frame area image, and extracts the upper lip feature point O and the lower lip feature point Q from it;

[0084] D2. Calculate the distance between the upper lip feature point and the lower lip feature point in each frame to obtain a distance change peak value. When the distance change peak value is greater than a preset peak value, it is determined that the user is speaking.

[0085] Wherein step D1 includes the following sub-steps:

[0086] D11, obtaining facial key points in the face frame area image, including the upper midpoint O1 of the upper lip, the lower midpoint O2 of the upper lip, the upper midpoint Q1 of the lower lip, and the lower midpoint Q2 of the lower lip;

[0087] D12. Extract the midpoint between the upper midpoint O1 of the upper lip and the lower midpoint O2 of the upper lip as the upper lip feature point O, and extract the midpoint between the upper midpoint Q1 of the lower lip and the lower midpoint Q2 of the lower lip as the lower lip feature point Q.

[0088] Among them, the specific method for obtaining the key points of the face in the face frame area image is as follows: Figure 3 As shown, first extract the eyebrow center K, nose center K, and chin center L in the face area, and use the three as anchor points to establish the contour center line I, and then obtain: the upper lip midpoint O1 and the lower lip midpoint O2 are the intersection points of the contour center line I and the upper edge of the upper lip, and the intersection points of the contour center line I and the lower edge of the upper lip, respectively, the lower lip upper midpoint Q1 and the lower lip midpoint Q2 are the intersection points of the contour center line I and the upper edge of the lower lip, and the intersection points of the contour center line I and the lower edge of the lower lip, respectively, and the upper lip feature point O is the midpoint of the line connecting the upper lip midpoint O1 and the lower lip midpoint O2, and the lower lip feature point Q is the midpoint of the line connecting the upper lip midpoint Q1 and the lower lip midpoint Q2. The above calculation and analysis method can effectively save computing power resources and improve computing efficiency.

[0089] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for managing user biometric information based on an artificial intelligence-based medical guidance system, characterized in that: The invention comprises a server (1) and an intelligent medical guidance device (2) which is communicatively connected to the server (1); the intelligent medical guidance device (2) is provided with an intelligent answering system for conducting voice dialogue with a user, and a display module (21) for displaying an AI virtual character and text and picture information; the intelligent medical guidance device (2) is also provided with a camera device (22) for collecting user image information, a sound receiving device (23) for collecting user sound information, and a sound generating device (24) for sound output; The method for managing user biometric information by the artificial intelligence-based medical guidance system comprises the following steps: S1, a user interacts with the intelligent medical guidance device (2), the user can register and establish a user account in the intelligent answering system, the system establishes a corresponding user database for the user account and uploads it to the server (1), the intelligent medical guidance device (2) collects the user's audio information through the sound receiving device (23), collects the user's facial image information through the camera device (22), and uses the collected facial image information as authentication information to identify the user, marks the user who meets the authentication information as a registered user, and binds the collected audio information and facial image information to the user account of the user and stores them in the corresponding user database; S2. When the user uses the intelligent medical guidance device (2) again and interacts with the user, the intelligent medical guidance device (2) compares the facial image information of the user with the facial image information in the user database and finds that the information is consistent. Then, the intelligent medical guidance device (2) calls the data in the user database of the registered user from the server (1), and the intelligent medical guidance device (2) interacts with the registered user normally, while collecting the current audio information of the registered user in real time. S3. In step S2, the intelligent medical guidance device (2) synchronously compares the current audio information of the registered user with the audio information in the user database. When the deviation value between the current audio information of the registered user and the audio information in the user database is large, the intelligent medical guidance device (2) actively sends a reminder message.

2. The method for managing user biometric information by an artificial intelligence-based medical guidance system according to claim 1, characterized in that: In step S1, the intelligent medical guidance device (2) extracts the initial pitch value fo, the initial loudness value L0 and the initial formant characteristic value F0 in the audio information corresponding to the user account; In step S3, after the intelligent medical guidance device (2) collects the audio information of the user, it extracts the current pitch value f, the current loudness value Ln and the current formant characteristic value F in the audio information; In step S3, the pitch difference fx between the current pitch value f and the initial pitch value fo, the loudness difference Lx between the current loudness value Ln and the initial loudness value L0, and the peak feature difference Fx between the current resonance peak eigenvalue F and the initial resonance peak eigenvalue F0 are calculated respectively, and it is determined that the deviation between the current audio information of the registered user and the audio information in the user database is large when the following conditions are met: the pitch difference fx is greater than or less than the preset range of the pitch difference, and the loudness difference Lx is greater than or less than the preset range of the loudness difference, and the peak feature difference Fx is greater than or less than the preset range of the peak feature difference.

3. The method for managing user biometric information based on an artificial intelligence-based medical guidance system according to claim 1, characterized in that: The steps of the intelligent medical guidance device (2) identifying and collecting the user's audio information include: A. When a user approaches the intelligent medical guidance device (2) and is recognized by the camera device (22), the intelligent medical guidance device (2) recognizes and calculates the real-time distance between the user and the device. When the real-time distance is less than or equal to a first preset distance, the intelligent medical guidance device (2) collects facial image information of the user; B. The intelligent medical guidance device (2) identifies and determines whether the user is wearing a mask. If the user is wearing a mask, the process proceeds to step C; if the user is not wearing a mask, the process proceeds to step D; C. When the user wears a mask, the intelligent medical guidance device (2) calculates the real-time distance between the user and the device. When the real-time distance is less than or equal to the second preset distance, the sound receiving device (23) starts to work and collects the user's voice information; D. When the user is not wearing a mask, the intelligent medical guidance device (2) calculates the user's lip movements through real-time facial image information to determine whether the user is speaking. When the user starts speaking, the sound receiving device (23) starts working and collects the user's voice information.

4. The method for managing user biometric information based on an artificial intelligence-based medical guidance system according to claim 2, characterized in that: The steps of extracting the pitch in the audio information include: A. performing frame processing on the audio information to obtain a plurality of framed audios; B. Performing Fourier transform on each frame of the divided audio in step A to obtain frequency domain data, obtaining the peak point frequency corresponding to the peak point in the frequency domain data, and determining the average value of the peak point frequency as the fundamental frequency; C. Based on the fundamental frequency in the framed audio, the pitch of the framed audio is calculated; D. Generate a pitch value of the audio information based on the pitch of the framed audio.

5. The method for managing user biometric information based on an artificial intelligence-based medical guidance system according to claim 2, characterized in that: The steps of extracting loudness from audio information include: A. The sound receiving device (23) captures the sound, converts it into an electrical signal and amplifies the electrical signal; B. The intelligent diagnosis guidance device (2) calculates the effective value of the electrical signal to obtain the average sound pressure P; C. According to the formula SPL=log10(P / P0), where P0 is the reference sound pressure, for the sound pressure in the air, P0 is 20 microPascals (20μPa), and the calculated SPL is the loudness value.

6. The method for managing user biometric information by an artificial intelligence-based medical guidance system according to claim 2, characterized in that: The steps of extracting the formant feature value in the audio information include: A. Acquire an audio signal, perform frame processing on the audio signal, and obtain N voice frames; B. Perform fast Fourier transform on each speech frame to obtain the spectrum of each speech frame, obtain the formant data of each speech frame based on the spectrum, and take the first three formant data of each speech frame; C. Calculate the statistical average of the three vibration peak data, which is the resonance peak characteristic value.

7. The method for managing user biometric information by an artificial intelligence-based medical guidance system according to claim 3, characterized in that: The steps of the intelligent medical guidance device (2) determining whether the user is speaking include: D1, a camera device (22) collects facial image information of a user, converts the facial image information into an image sequence, extracts a face region of each frame in the image sequence, identifies the face part in the face image based on a convolutional neural network (CNN) architecture, and then obtains face key points in the face frame region image, from which upper lip feature points O and lower lip feature points Q are extracted; D2. Calculate the distance between the upper lip feature point and the lower lip feature point in each frame to obtain a distance change peak value. When the distance change peak value is greater than a preset peak value, it is determined that the user is speaking.

8. The method for managing user biometric information based on an artificial intelligence-based medical guidance system according to claim 7, characterized in that: Step D1 includes the following steps: D11, obtaining facial key points in the face frame area image, including the upper midpoint O1 of the upper lip, the lower midpoint O2 of the upper lip, the upper midpoint Q1 of the lower lip, and the lower midpoint Q2 of the lower lip; D12. Extract the midpoint between the upper midpoint O1 of the upper lip and the lower midpoint O2 of the upper lip as the upper lip feature point O, and extract the midpoint between the upper midpoint Q1 of the lower lip and the lower midpoint Q2 of the lower lip as the lower lip feature point Q.

Images