classifying and grouping the second audio information based on the similarity to obtain a plurality of audio groups; dividing the audio signals with the similarity larger than a judgment threshold (for example: 0.90) into an audio group;

sorting the second audio information based on the acquisition time;

when the time difference value is less than or equal to a preset first time threshold (for example: 2 minutes), the audio group is not split; otherwise, splitting the audio group between two pieces of second audio information of which the time difference value is greater than the first time threshold; through splitting processing, each audio group is ensured to be independent continuous audio, so that the audio of two adjacent patients behind can be divided;

determining a first time range corresponding to the audio group based on the acquisition time corresponding to the first second audio information of the audio group and the acquisition time corresponding to the last second audio information;

when the first time range corresponding to the audio group is not overlapped with the first time ranges corresponding to other audio groups, taking the first audio information corresponding to the first second audio information of the audio group as a starting position and the last second audio information of the audio group as an ending position, cutting the second audio and video, and acquiring a third audio and video;

when the ratio of the overlapping time to the maximum time difference is smaller than or equal to a preset ratio (for example: 0.1), taking the first audio information corresponding to the first second audio information of the audio group as a starting position and the last second audio information of the audio group as an ending position, clipping the second audio and video, and acquiring a third audio and video;

the audio of the second audio information is divided into each audio group by extracting the characteristics of the audio, so that the audio of the same person is divided into one group; normally a person's audio represents a patient; however, there is a case where a patient does not visit himself or herself; for example: pediatrics, often accompanied by father and mother; thus by whether the audio groups overlap, and the ratio of the overlap; determining whether the person is accompanied by a plurality of persons; the accurate determination of the starting position and the ending position of the third audio is realized, and the accuracy of audio editing is improved.

In one embodiment, thevideo analysis module 2 performs grouping and aggregation on the plurality of third audios and videos based on the first audio feature set and the human body contour map to obtain a plurality of fourth audios and videos, and performs the following operations:

when the human body contour map or the first audio feature set is matched, two third audio-video sets which are matched with each other are integrated.

the body contour map and the feature set of the audio frequency in the third audio-video frequency are comprehensively analyzed, so that the integration of the third audio-video frequency when the patient enters for the first time and enters for the second time is realized, and the integrity of the inquired audio-video frequency is ensured.

In one embodiment, theassociation module 3 associates the inquiry videos with the patient medical records one by one, stores and forms the patient information, and performs the following operations:

acquiring operation information of a computer in an interrogation room;

analyzing the operation information, and determining the calling time of the registration information of each patient; the time that the registration information of the patient, which is clicked by the doctor and enters the diagnosis page, is determined as the calling time;

determining a second time range for the interrogation of each patient based on the call time; the calling time of the registration information of the previous patient to the calling time of the registration information of the next patient can be used as a second time range for the inquiry of one patient;

acquiring a time range set of inquiry audios and videos; the set of time ranges includes at least one third time range; when the patient enters the consulting room for multiple times, multiple third time ranges exist in the time range set;

determining a correspondence of the interrogation audio video to the patient medical records based on the second time range and a plurality of third time ranges; when the first third time range and the second time range have an overlap of a preset proportion (50%), it can be determined as corresponding; for example: the second time range falls within a first third time range.

In one embodiment, the request receiving module 4 receives a medical record acquisition request, as shown in fig. 2, and performs the following operations:

step S1: acquiring the actual position of a requester of the medical record acquisition request;

step S2: acquiring a preset permission area for permitting to receive a request; the licensed region may be defined as within a hospital;

step S3: when the actual position belongs to the permission area, receiving a medical record acquisition request; otherwise, it does not receive.

And legality judgment of the request is carried out according to the actual position of the request, so that the safety of the data recorded by the medical record of the patient is ensured.

In one embodiment, theverification module 5 verifies the medical record acquisition request and performs the following operations:

based on the contact information (such as the mobile phone number) in the reserved information, sending a verification code;

and when the verification code is received from the requester of the medical record acquisition request within a preset second time threshold (for example, 1 minute), the verification is passed.

by sending the verification code to the reserved mobile phone for verification, the data of the medical record acquired under the authorization of the patient is ensured, and the privacy of the patient is effectively protected.

In one embodiment, the AI and trusted based medical cloud communication system, as shown in fig. 3, further comprises:

thedata acquisition modules 8 are correspondingly in communication connection with the detection instruments in the hospital and used for acquiring detected original data;

and theassociation module 3 is further used for associating the original data with the patient medical record and storing the original data and the patient medical record to form the detection data of the patient.

through correlating the detected original data with the patient medical record, the hospital receiving the patient can acquire the original data under the conditions of patient transfer and the like, and the occurrence of misdiagnosis accidents can be further avoided.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims

1. A medical cloud communication system based on AI and innovation, comprising:

the system comprises a video acquisition module, a first audio and video acquisition module and a second audio and video acquisition module, wherein the video acquisition module is used for acquiring a first audio and video in an interrogation room, which is shot by audio and video acquisition equipment arranged in the interrogation room;

the information acquisition module is used for acquiring the patient treatment information corresponding to the medical record acquisition request when the medical record acquisition request passes verification;

the transmission module is used for transmitting the acquired patient treatment information to a requester of the medical record acquisition request;

the video analysis module analyzes and clips the first audio and video based on artificial intelligence to obtain a plurality of inquiry audios and videos, and executes the following operations:

based on the number of human body contours existing in each frame of image, clipping the first audio/video to obtain a plurality of second audio/video;

extracting a human body contour map at a preset first position in an image in the third audio-video frequency;

extracting the characteristics of the audio corresponding to the second audio information interacted with the inquiry doctor in the third audio-video to obtain a first audio characteristic set;

grouping and aggregating a plurality of third audios and videos based on the first audio feature set and the human body contour map to obtain a plurality of fourth audios and videos;

and taking the fourth audio-video as the inquiry audio-video.

2. The AI and creative-based medical cloud communication system of claim 1, wherein said video analysis module clips said first audio-video based on the number of human body contours present in each frame of image, obtaining a plurality of second audio-video, performing the following operations:

deleting the frame of image when only one human body contour exists or the human body contour does not exist in the image;

and integrating the continuous frame images into the second audio-video.

3. The AI-and-trauma-based medical cloud communication system of claim 1, wherein said video analysis module identifies audio in each of said second audio videos, determines first audio information of an inquiring physician and second audio information of each of said second audio videos interacting with said inquiring physician, and performs the following operations:

respectively extracting keywords of the audio information to be processed based on a preset keyword extraction template to obtain keyword extraction results;

taking the query information as the first audio information and the answer information as the second audio information;

and/or the presence of a gas in the atmosphere,

respectively extracting the characteristics of the audio corresponding to the audio information to be processed to obtain a plurality of characteristic values;

inquiring a preset identification factor and personnel comparison table based on the identification factor, and determining that the audio information to be processed corresponds to an inquiry doctor in the identification factor and personnel comparison table or corresponds to other personnel in the identification factor and personnel comparison table;

and taking the audio information to be processed corresponding to the doctor of the inquiry as first audio information, and taking the audio information to be processed corresponding to the other personnel as second audio information.

4. The AI and trusted based medical cloud communication system of claim 1, wherein said video analytics module clips said second audio video based on said second audio information, obtaining a plurality of third audio video, performing the following operations:

calculating the similarity between the second audio feature sets;

sorting the second audio information based on the acquisition time;

when the time difference value is less than or equal to a preset first time threshold value, the audio group is not split; otherwise, splitting the audio group from the two pieces of second audio information of which the time difference value is greater than the first time threshold value;

when the first time range corresponding to the audio group is not overlapped with the first time ranges corresponding to other audio groups, the first audio information corresponding to the first second audio information of the audio group is taken as a starting position, the last second audio information of the audio group is taken as an ending position, the second audio and video are clipped, and the third audio and video are obtained;

when the ratio of the overlapping time to the maximum time difference is smaller than or equal to a preset ratio, taking first audio information corresponding to first second audio information of the audio group as a starting position and taking last second audio information of the audio group as an ending position, clipping the second audio and video, and acquiring a third audio and video;

and when the ratio of the overlapping time to the maximum time difference is greater than a preset ratio, combining the overlapped audio groups, taking the first audio information corresponding to the first second audio information of the combined audio groups as a starting position and the last second audio information of the combined audio groups as an ending position, clipping the second audio and video, and acquiring the third audio and video.

5. The AI and creative-based medical cloud communication system of claim 1, wherein said video analysis module groups and aggregates said third audios and videos based on said first set of audio features and said human body profile to obtain a fourth plurality of audios and videos and performs the following operations:

matching each of the first audio feature sets of the third audio and video with each of the first audio feature sets of the other third audio and video;

and when the human body contour map or the first audio feature set is matched, integrating the two matched third audio and video.

6. The AI-and-trauma-based medical cloud communication system of claim 1, wherein said association module associates said interrogation audio-video with patient medical records one-to-one, stores and forms patient visit information, performs the following operations:

acquiring operation information of a computer in the consulting room;

acquiring a time range set of the inquiry audio/video; the set of time ranges comprises at least one third time range;

determining a correspondence of the interview audio to the patient medical records based on the second time range and a plurality of the third time ranges;

and associating the inquiry audio-video with the patient medical record based on the corresponding relation.

7. The AI-and-based-medical cloud communication system of claim 1, wherein said request receiving module receives a medical record acquisition request by:

acquiring a preset permission area for permitting to receive a request;

when the actual position belongs to the permitted area, receiving the medical record acquisition request; otherwise, it does not receive.

8. The AI-and-based medical cloud communication system of claim 1, wherein said verification module verifies said medical record acquisition request by:

when the medical record acquisition request is received, acquiring the reserved information of the patient;

sending a verification code based on the contact information in the reserved information;

9. The AI and innovation-based medical cloud communication system of claim 1, further comprising:

the data acquisition modules are in corresponding communication connection with all detection instruments in the hospital and used for acquiring detected original data;