Discrete artificial intelligence diagnosis and treatment system

Abstract

The biggest feature of this system is that it uses a discrete structure to transform the manual operation mode of one doctor and one patient into a systematic operation mode, and the two steps of diagnosis and treatment are handled by two systems respectively, namely the customer system and the central system. Using artificial intelligence, the procedures of looking, smelling, asking and palpating are completed, the disease data is detected and converted into a combination of numbers and sent to the central system for processing. The system has a powerful memory function. It can record more than one million customer files and tens of thousands of one-disease-multiple-prescription combinations, including one main prescription and several auxiliary prescriptions. Each prescription combination is crowned with a number combination format code. When a customer sends a number combination, the central system will find an identical number combination in the memory library, which is attached with the corresponding prescription combination. The system will select a prescription suitable for the patient's constitution based on the patient's file. The system has high-speed data processing capabilities and can process tens of thousands of cases every day, greatly improving medical work efficiency.

Description

Technical Field

[0001] This is an invention belonging to the field of medical technology, which aims to solve the social problems of expensive and difficult medical treatment for the people and save medical expenses for the country. Background Art

[0002] Since ancient times, medical technology has been operated manually, but with the development of technology, some new medical methods have appeared on the market:

[0003] Both medical consultation and telemedicine are Internet+ medical services, but they are one-to-one services and cannot meet the medical needs of large numbers. There are also robot doctors on the market. Its basic model is still a robot replacing a doctor. The price is very high, and the technology is not mature yet. It can only be regarded as a transitional product and it is easy to be eliminated due to upgrading. The above services and products cannot meet the medical needs of large numbers and solve the existing problems of expensive and difficult medical treatment. Summary of the invention

[0004] The present invention aims to create a medical system that can solve the society's eager demand for medical problems. The discrete artificial intelligence diagnosis and treatment system is targeted at the above problems because such a system can provide a large number of medical service resources, has low operating costs, and can be replicated indefinitely.

[0005] Our approach is to use modern technology and traditional medicine to organically combine, using technology to greatly improve the efficiency of medical work, with the goal of allowing the people to receive high-quality, cheap and convenient medical services. DETAILED DESCRIPTION

[0006] The manual operation model of one doctor and one patient that we currently use has been used for more than 2,000 years. This is an inefficient model that is completely out of touch with modern technology. We must transition to an intelligent and systematic model to effectively provide high-quality medical services in large quantities at low cost.

[0007] We organically combine existing medical technology with modern science and technology, and use scientific and technological methods to improve the efficiency of medical work. The specific approach is to introduce a discrete artificial intelligence diagnosis and treatment system. We handle the two steps of diagnosis and treatment separately. The "diagnosis" is handled by the customer system, which is composed of hundreds of thousands of artificial intelligence diagnostic devices; the "treatment" is handled by the central system, which is actually a large computer that can record more than one million patient files, including personal information, medical records, physical examination reports, allergy history, etc. It also records tens of thousands of supporting materials for multiple prescriptions for one disease.

[0008] With the help of artificial intelligence tools, people can self-diagnose at home when they are sick. The client system converts the data obtained from the "look, smell, ask, and touch" program into a series format and sends it to the central system via the Internet. It has high-speed data processing capabilities and can complete the diagnosis, prescribe medicine and quote back to the client within one second. The system can handle tens of thousands of cases a day and match more than one million customers into a huge medical system (because the number of people who get sick every day is only about 10,000). This is a good division of labor. We let the client handle those scattered and time-consuming tasks, and then send the obtained data to the central system via the Internet for centralized and high-speed processing.

[0009] Next, let's talk about how the central system diagnoses and prescribes medicine. After more than 2,000 years of practice, our ancestors have basically mastered the treatment methods for each disease. In other words, each disease has one or more treatments, which is equivalent to a format. The results of "looking, smelling, asking, and touching" will display various diseases. We convert these results into a sequence format to make it easier for computers to process. That is to say, a set of sequence combinations represents a certain disease. We put together a main prescription for treating this disease and several auxiliary prescriptions derived from this prescription to form a set of combinations including sequence combinations, diseases and prescriptions, which are stored in the central system. Among them, the main prescription is applicable to most patients, but the patients are diverse, including pregnant women, patients with underlying diseases, allergic to different drugs, etc. The main prescription may not be suitable for these patients, so some auxiliary prescriptions need to be added for selection in order to meet the requirements of personalized prescriptions.

[0010] When the central system receives the number combination sent by the patient, it will automatically find an identical number combination in the memory bank, that is, find the matching prescription combination corresponding to the condition, then open the patient's file to read the information, and select the most suitable prescription based on the patient's personal situation.

[0011] Because each has its own method, there may be several main prescriptions for treating a disease. Therefore, when we implement it, we need a large number of senior doctors to write multiple main prescriptions and auxiliary prescriptions, and these prescriptions must be verified. We use a system to verify a main prescription and its auxiliary prescriptions. We use telephone follow-up to track the effect of each prescription, eliminate the inefficient ones, and only the efficient ones are left, so that all systems can be used together. Only a large system like ours can conduct large-scale screening and screening of different prescriptions left by the people to determine their effectiveness. This is a valuable and huge project.

[0012] We will sell AI diagnostic devices to the public at affordable prices, and those who do not want to buy these diagnostic devices can go to our service stations for diagnosis and treatment. There are a large number of equipment, customer service and doctors provide 24-hour service.

[0013] This system currently only focuses on some non-fatal diseases and non-surgical patients. If the patient's condition has not improved after using the medicine we provide for a few days, or the patient sends the information to the central system and cannot prescribe the medicine, the customer service will recommend the patient to go to the service station for diagnosis and treatment. This is an arrangement to solve special situations.

[0014] This system frees a large number of doctors from their daily tedious work. Some people may resist it, but it is a highly anticipated project and an inevitable trend of the times. No one can stop it. But it is certain that the government will take good care of the lives of doctors and arrange new jobs for them.

[0015] The discrete artificial intelligence diagnosis and treatment system has unparalleled advantages. It can help us transition from the one-doctor-one-patient operation mode to an intelligent and systematic mode. This will be a leap forward and will lead other countries in the world to carry out similar medical reforms.

[0016] Next, let's talk about how we convert the data obtained by "look, smell, ask, and touch" into a number series format. First, we use "1, 2, 3, 4" as the codes for "look, smell, ask, and touch" respectively, and these codes will be placed in the first position of each number series.

[0017] The main content of the observation is the observation of the oral cavity. We use the method of taking photos to complete the scope of observation, including 1. oral wall 2. uvula 3. tongue 4. gums, etc. The main symptoms of illness are 1. inflammation 2. ulcer 3. white fur on the tongue surface 4. yellow fur on the tongue surface, etc. The numbers representing different positions of the oral cavity are placed in the second place of the sequence, and the numbers representing different symptoms are placed in the third place of the sequence.

[0018] Example 1: The first number in the sequence 1, 2, 1 is 1, which means this is the sequence of vision, the second number is 2, which means this is the location of the uvula, and the third number is 1, which means there are symptoms of inflammation.

[0019] Example 2: The first number in the sequence 1, 1, 2 is 1, which means it is a sequence of vision; the second number is 1, which means it is the location of the oral wall; and the third number is 2, which means there are symptoms of ulcers.

[0020] After taking the photo, the AI ​​software will automatically generate a sequence of numbers that we will operate using our smartphones.

[0021] The code for smell is 2, which is auditory perception, mainly to understand the coughing situation, but coughing may not occur during self-examination, so here we use the questioning method instead. But the questioning here is only to understand the coughing situation, which is different from the "ask" later. Open the software, after the smell option, there will be a series of options for the patient to click to complete.

[0022] The coughing time is: 1. Daytime 2. Nighttime 3. Day and night 4. On an empty stomach. These data will be placed in the second position of the sequence. The degree of coughing is: 1. Occasionally 2. Paroxysmal cough 3. Long cough. These data will be placed in the third position of the sequence. The sputum conditions are: 1. Dry cough without sputum 2. White with bubbles 3. Yellow... These data will be placed in the fourth position of the sequence.

[0023] Example: 2, 2, 3, 1 means this is the number sequence of "smell", which means long, dry cough without phlegm at night.

[0024] The code for "ask" is 3, which means "inquiry". We use the mobile phone to display a series of options for patients to click on: 1. Dizziness 2. Diarrhea 3. Vomiting 4. Constipation 5. Loss of appetite 6. Tinnitus 7. Runny nose 8. General fatigue 9. Decreased vision 10. Partial body pain... The above options will be placed in the second position of the sequence.

[0025] When you click on the options 1-9, you can select one or more items; if you click on option 10, the screen will show a full-body image of a human body, with the front and back sides allowing the patient to click on specific locations on the body. The third number in the sequence represents the location of the pain. The fourth number in the sequence represents the degree of pain, which can be divided into multiple levels.

[0026] For example: 3, 2; 3, 3; 3, 10, a, 2 means this is a "question" sequence. The patient has diarrhea, vomiting, and pain in the body position a.

[0027] The code for pulse diagnosis is 4. There are 28 pulse diagnosis waveforms, i.e. 1-28. We store these 28 waveforms in an electronic device. Use a pulse diagnosis instrument to read the patient's pulse waveform, input these waveforms into the device that stores the 28 waveforms, and then use artificial intelligence to compare the waveform just input to find out which waveform is most similar to the input waveform, so as to obtain the category code of the input waveform. This sequence is 4a, where 4 is the pulse diagnosis code and a is the pulse diagnosis classification, ranging from 1-28. The sequence of "look, smell, ask" may be one or more, but the sequence of "diagnosis" can only be one.

[0028] The above content is only used to explain the present invention and is not authoritative. The actual content will be determined by doctors.

[0029] So far, I have talked about how to use the series method to express the information obtained by artificial intelligence through looking, smelling, asking and touching.

[0030] Let me make a summary below: First of all, we should realize that diagnosis and treatment are formatted, not disordered, but there may be multiple prescriptions to treat each disease, and there are also multiple physical conditions of patients. We must conduct large-scale experiments to determine the effectiveness of these prescriptions, eliminate the dross and retain the essence, and only retain those high-quality prescriptions for use. This will improve the treatment effect of this system and allow Chinese medicine to be carried forward.

[0031] The discrete structure is a good arrangement. It is a scattered and time-consuming task to obtain medical information from different patients. We leave these tasks to the client and convert the medical information into a series format, which is then sent to the central system via the Internet for rapid processing.

[0032] This system allows national doctors to incorporate their lifelong learning into it and provide services to the people now and a hundred years from now. This is a great honor. After this system matures, we only need to copy about 1,000 sets to meet the needs of the people across the country.

[0033] This is a project that can make the medical industry leap forward at a negligible cost. The government can save a lot of medical expenses to support the development of other fields. If state-owned enterprises and listed companies are allowed to operate, most of the profits will return to the country in the future, which can also show that the country's prescription policy is real. At the same time, it can also stimulate the creativity of the people and trigger more projects of great significance. I implore the Intellectual Property Office to announce this project as soon as possible so that investors can participate in the investment as soon as possible.

Claims

1. The feature of the present invention is a discrete structure. The entire system is divided into two major parts: a central system and a client system. They have a clear division of labor. The client system is composed of a large number of intelligent detectors. The function of these detectors is to detect the patient's medical information and then send this information to the central system through the Internet so that this information can be processed centrally.

2. The second feature is that the data collected through the customer's system of looking, smelling, asking and touching will be expressed in the form of a series, and each combination of the series corresponds to a disease.

3. The third feature is that the central system stores tens of thousands of prescription combinations written by traditional Chinese medicine practitioners. Each combination contains a main prescription and several auxiliary prescriptions. The combination is preceded by a number series combination as a serial number, which corresponds to a disease. Here, the serial number, disease, and prescription form a large combination.

4. The fourth feature is that when the central system receives a number sequence combination sent by a patient, it will search the prescription library for a number sequence combination that is exactly the same as the combination just received, thereby obtaining the large combination mentioned in claim 3, which contains the prescription combination required by the visiting customer.

5. The fifth feature is that the system needs to open the file of the visiting patient to refer to the contents therein, and then find a suitable prescription from the prescription combination obtained in claim 4.

6. The sixth feature is the method of constructing a combination of number series. We use 1, 2, 3, and 4 as the codes for look, smell, ask, and cut respectively. These codes will be placed in the first position of the number series, and other variables will be placed in the second, third, fourth, etc. positions (if any).

7. The seventh feature is the "ask" approach, which provides a table through the mobile phone, showing various conditions for patients to click according to the actual situation. If the patient has a pain option, the phone will display the front and back.

8. The eighth feature is the visual processing method. We use a smartphone to take a photo of the oral cavity, and then use artificial intelligence to convert the photo results into one or more number series.

9. The ninth feature is the confirmation of "cut". We pre-store 28 kinds of "cut" waveforms in an electronic device and number them from 1 to 28. Then we use a pulse cutting device to cut the patient's pulse waveform, and then compare the waveform with the waveform stored inside to find out which pre-stored waveform the input waveform is most similar to. In this way, we can confirm the classification of the input waveform, and then convert the result into a series format, such as 4a, where 4 is the code for "cut" and a is the number of the waveform, which is between 1 and 28.