Prevention reminding system for lower limb venous thrombosis

Abstract

The invention provides a prevention reminding system for lower limb venous thrombosis, which is based on a built-in lower limb venous thrombosis data model. According to symptoms input by a user, family thrombus history, past thrombus history, whether an operation is carried out or not and operation types, the historical process experienced by forming the current lower limb venous thrombosis state, the development process of the lower limb venous thrombosis of the patient according to the current state and the life quality of the lower limb venous thrombosis patient based on a recommended treatment method are displayed to the patient; the prevention reminder system uses a display or a tablet computer when shown to a patient. According to the invention, the patient can accurately understand the diagnosis of the doctor, can clearly understand the reason of the current illness state, and can actively cooperate with the treatment of the doctor, the misunderstanding between the doctor and the patient is avoided, and the barrier of the traditional outpatient diagnosis and treatment is broken through.

Description

Technical Field

[0001] The invention belongs to a prevention reminder system, and in particular relates to a prevention reminder system for lower limb venous thrombosis. Background Art

[0002] Lower extremity venous thrombosis includes lower extremity superficial venous thrombosis and lower extremity deep venous thrombosis. Although the clinical symptoms of lower extremity superficial venous thrombosis are mild, if it is not treated promptly and effectively, it will develop into lower extremity deep venous thrombosis. Lower extremity deep venous thrombosis often manifests externally as swelling, pain, temperature rise, color change, heaviness of the lower extremities, and abnormal sensation. If it is not treated, it will develop into lower extremity deep vein thrombosis syndrome, and more seriously, it will cause pulmonary embolism.

[0003] Therefore, when patients encounter deep vein thrombosis in the lower limbs, they need to seek medical treatment in a timely manner, otherwise the condition will worsen further. In severe cases, they will lose the ability to move, affecting the patient's quality of life and increasing the burden on the family.

[0004] However, due to the short time of outpatient clinics, patients generally cannot have a clear understanding of the severity of their illness through the doctor's oral description in a short period of time. Not only can they not cooperate with the doctor's treatment, but it also increases the psychological burden on patients.

[0005] Not only do we encounter the above problems in actual work, but similar phenomena also exist among other colleagues. In the existing literature, "Nursing Experience in Thrombolytic Treatment of Deep Vein Thrombosis of Lower Limbs", Chen Xiaomin, "Guangxi Medicine", P1819-1820; "Health Education for the Prevention of Deep Vein Thrombosis in Patients with Lower Limb Fractures", Zhang Yi, "Medical Information", P259-P260; "Chief Physician Lei Xiaoming's Discussion and Treatment Experience on the Classification of Deep Vein Thrombosis of Lower Limbs", Li Zhengang et al., "Journal of Hebei Traditional Chinese Medicine", P42-P43; "Anticoagulant Treatment of Deep Vein Thrombosis of Lower Limbs after Hip Surgery", Qian Sulin et al., "Transportation Medicine", P177; it is mentioned that patients' lack of understanding of deep vein thrombosis of lower limbs has resulted in missing the best time for treatment.

[0006] Therefore, it is necessary to provide a solution that can improve patients' understanding of lower extremity venous thrombosis in a shorter period of time or in outpatient clinic time, while enabling patients to understand the treatment methods and treatment effects and improve the cooperation between doctors and patients. Summary of the invention

[0007] In view of the above technical problems, the present invention provides a prevention and reminder system for lower limb venous thrombosis, aiming to overcome the problems in the above-mentioned prior art and enable patients to have a comprehensive understanding of their current conditions in a longer period of time.

[0008] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0009] A prevention and reminder system for lower limb venous thrombosis, which is based on a built-in lower limb venous thrombosis data model. According to the symptoms, family history of thrombosis, previous history of thrombosis, whether surgery was performed and the type of surgery input by the user, the patient is shown the historical process of the current lower limb venous thrombosis state, the development process of the patient's lower limb venous thrombosis according to the current state, and the quality of life of the patient with lower limb venous thrombosis based on the recommended treatment method; the prevention and reminder system uses a display or a tablet computer to display to the patient.

[0010] Further,

[0011] The lower limb venous thrombosis data model is based on a database formed by cases received by the hospital and uses a CNN deep learning model to establish a generation model of the historical process, development progress, and quality of life of lower limb venous thrombosis in historical patients; the input data of the CNN deep learning model are the external manifestations, internal manifestations, and treatment methods of historical patients; the external manifestations include: pain, redness and swelling, skin temperature, and skin color; the internal manifestations include the location of the thrombus, the size of the thrombus, the morphology of the thrombus, and the blood filling degree; the treatment methods include not using treatment methods, the types of treatment methods used, and whether regular examinations are performed.

[0012] Further,

[0013] The prevention reminder system is also provided with a prevention model for lower limb venous thrombosis, which is established based on the data model of lower limb venous thrombosis and covers treatment methods for different stages of lower limb venous thrombosis; the prevention model generates multiple treatment methods based on the current condition of the patient's lower limb venous thrombosis, and sorts the multiple treatment methods based on the historical treatment effects of the patients, and displays them on a display screen or tablet computer as a recommendation to the patient; the patient can make a choice based on his or her own situation or according to the doctor's advice.

[0014] Further,

[0015] Based on the treatment conditions of new patients, the database is updated, and the CNN deep learning model is used to update the historical process, development process and quality of life model in the lower limb venous thrombosis data model. The prevention model is rebuilt according to the updated lower limb venous thrombosis model, and the treatment methods are updated; then the treatment methods and rankings are manually checked and corrected.

[0016] Further,

[0017] The treatment methods include physical therapy, drug therapy and surgical treatment. The prevention model also recommends specific treatment methods based on whether the patient has a risk of bleeding. The treatment methods can be single or a combination of multiple methods.

[0018] Further,

[0019] The prevention reminder system is also provided with a risk assessment model, which displays the external and internal manifestations of the lower limbs in stages based on the lower limb venous thrombosis data model and the treatment method selected by the patient received by the prevention model. The stages are divided into treatment stages or time stages, and the display time of each stage is set to 10 seconds. A total of 5 stages are set, and the patient makes a selection on the display or tablet computer according to the treatment effect to be achieved.

[0020] Further,

[0021] The historical process refers to the process of lower limb venous thrombosis that the patient has experienced before the current condition; the development process refers to the process of lower limb venous thrombosis that the patient will face if he continues his current lifestyle and treatment methods; the quality of life refers to the physical and psychological state of the patient with or without treatment methods; when establishing the database, the physical and psychological states are obtained through observation, examination and feedback from patients.

[0022] Further,

[0023] For the CNN deep learning model, the following formula is used to calculate the initial weight of the external performance:

[0024]

[0025] Among them, γ1 represents the pain level at the initial detection, τ1 represents the redness and swelling level at the initial detection, T1 represents the skin temperature at the initial detection, c1 represents the skin color level at the initial detection, λ1, λ2, λ3, λ4 are the parameters of pain, redness and swelling, skin temperature and skin color respectively;

[0026] The initial weights are corrected, and pain, redness, swelling, skin temperature and skin color are detected multiple times at set intervals, which are recorded as: γ i , τ i , T i 、c i , i = 1, 2, ... n, calculate the difference of the above detection objects. If the average value of the difference is less than the set threshold, the initial weight is used as the weight of the external performance in the CNN deep learning model; as long as the average value of the difference of one of the detection objects is greater than the set threshold, the initial weight needs to be corrected. The calculation formula is:

[0027]

[0028] Among them, γ represents the average pain level during detection, τ represents the average redness and swelling level during detection, T represents the average skin temperature during detection, c represents the average skin color level during detection, and Δ is the product of the corresponding parameter and the corresponding difference average value.

[0029] Further,

[0030] For the CNN deep learning model, the following formula is used to calculate the weight of the intrinsic representation:

[0031]

[0032] Among them, p represents the influencing factor of thrombus position, s represents the influencing factor of thrombus size, η represents the influencing factor of thrombus morphology, ρ represents the blood filling degree, κ1 represents the parameter of the influencing factor of thrombus position, κ2 represents the parameter of the influencing factor of thrombus size, and κ3 represents the parameter of the influencing factor of thrombus morphology.

[0033] Further,

[0034] The weights of each external manifestation are: pain weight is λ1O z , redness weight is λ2O z , the skin temperature weight is λ3O z , the skin color weight is λ4O z ;

[0035] The weights of each intrinsic performance are as follows: the weight of thrombus location is κ1I, the weight of thrombus size is

[0036] κ2I, thrombus morphology weight is κ3I;

[0037] And according to each weight and external and internal manifestations, treatment methods are presented;

[0038] Or, if there are multiple treatment methods, set the external and internal floating weight influence factors F o and F I ,

[0039]

[0040] Then change the weight,

[0041] The weights of each external manifestation are: The redness weight is λ2O z , the skin temperature weight is λ3O z , the color weight of the skin is

[0042]

[0043] The weights of each intrinsic performance are as follows: The weight of thrombus location is The weight of thrombus size is The weight of thrombus morphology was κ3I;

[0044] And according to the weights and external and internal manifestations, treatment methods are given.

[0045] The beneficial effects of the present invention are as follows: The purpose of the present invention is to provide a visual outpatient device for lower limb venous thrombosis clinics that can be selected and reminded according to the different conditions of patients, so that doctors can have a more appropriate understanding of the doctor's diagnosis and subsequent treatment through the device or system on the basis of traditional medical consultation and explanation. Not only can patients accurately understand the doctor's diagnosis, but also patients can clearly understand the causes of the current condition, and can actively cooperate with the doctor's treatment. There is no misunderstanding between doctors and patients, breaking the barriers of traditional outpatient diagnosis and treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a flow chart of the use of the prevention reminder system of the present invention; DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0048] The present invention provides a lower limb venous thrombosis prevention reminder device or system for outpatient use. The device or system is a computer or tablet computer, or is mounted on a computer or tablet computer. It has a display or screen that can be touched or selected or input by a mouse, for lower limb venous thrombosis patients to watch and operate, so that they can clearly understand their own condition and provide a basis for cooperating with the doctor's treatment.

[0049] The lower limb venous thrombosis prevention reminder device or system of the present invention includes a lower limb venous thrombosis data model, a prevention model and a risk assessment model, and serves as an auxiliary device for outpatient consultation.

[0050] The lower limb venous thrombosis data model is a database established based on the cases received by the hospital, or not just the cases received by one hospital. It can be a database established by accumulating patient cases in sister hospitals or regions over a certain period of time. The more data collected, the better, and the more comprehensive the data collection for a single patient, the better.

[0051] The contents of the database records can be hospital examination data, diagnosis reports, patient inquiry records, etc., and should be collected as detailed as possible. When establishing database records, an incomplete case can be discarded, and cases with complete diagnosis and treatment content can be selected as much as possible, which will help with later data analysis. The database contains the patient's age, weight, tumor condition, surgical condition, previous medical history, living habits, injuries, chronic diseases, etc. It also includes the patient's symptoms of lower extremity venous thrombosis, the treatment methods used, and recovery status.

[0052] In the lower limb venous thrombosis data model, a deep learning model is used to establish a generative model of the historical process, development progress, and quality of life of lower limb venous thrombosis in historical patients. Available options include deep neural networks DNN, convolutional neural networks CNN, recurrent neural networks RNN, etc., or a combination of multiple deep learning models.

[0053] The input data of this deep learning model are the external manifestations of historical patients (obtained through the patient's feelings and diagnosis), internal manifestations (obtained through diagnostic examinations) and treatment methods (required when building the model). The most appropriate treatment method is calculated based on the weights and external and internal manifestations.

[0054] External manifestations include: pain, redness, swelling, skin temperature and skin color, etc.

[0055] Intrinsic manifestations include the location of the thrombus, the size of the thrombus, the morphology of the thrombus and the blood filling degree.

[0056] Treatment methods include no treatment, type of treatment used, and whether regular checkups are performed.

[0057] In the present invention, the specific deep learning model adopts the model in the prior art, which is not elaborated in the present invention. The model can be selected or combined according to the advantages and disadvantages of various deep learning models summarized in the prior art.

[0058] In an achievable embodiment, a CNN deep learning model is used. Different from other classifications, when a CNN deep learning model is used, the most important thing for the present invention is to adjust the weight of the input data, which is also the most critical part of the present invention. Therefore, the following is an appropriate setting of the weight of the input data so as to meet the judgment effect to be obtained by the present invention.

[0059] In one embodiment of the present invention, the following formula is used to calculate the initial weight of the external performance:

[0060]

[0061] Among them, γ1 represents the pain level at the initial detection, τ1 represents the redness and swelling level at the initial detection, T1 represents the skin temperature at the initial detection, c1 represents the skin color level at the initial detection, λ1, λ2, λ3, and λ4 are parameters of pain, redness and swelling, skin temperature, and skin color, respectively.

[0062] In order to make the deep learning model more suitable for the model of lower extremity venous thrombosis, the initial weights were corrected, and pain, redness, swelling, skin temperature and skin color were detected multiple times at set intervals, which were recorded as: γ i , τ i , T i 、c i , i = 1, 2, ... n, calculate the difference of the above detection objects. If the average value of the difference is less than the set threshold, the initial weight is used as the weight of the external performance in the CNN deep learning model; as long as the average value of the difference of one of the detection objects is greater than the set threshold, the initial weight needs to be corrected. The calculation formula is:

[0063]

[0064] Among them, γ represents the average pain level during detection, τ represents the average redness and swelling level during detection, T represents the average skin temperature during detection, c represents the average skin color level during detection, and Δ is the product of the corresponding parameter and the corresponding difference average value.

[0065] The preferred solution is to use the following formula to calculate the weight of the intrinsic performance:

[0066]

[0067] Among them, p represents the influencing factor of thrombus position, s represents the influencing factor of thrombus size, η represents the influencing factor of thrombus morphology, ρ represents the blood filling degree, κ1 represents the parameter of the influencing factor of thrombus position, κ2 represents the parameter of the influencing factor of thrombus size, and κ3 represents the parameter of the influencing factor of thrombus morphology.

[0068] The weights of each external manifestation are: pain weight is λ1O z , redness weight is λ2O z , the skin temperature weight is λ3O z , the skin color weight is λ4O z ;

[0069] The weights of each intrinsic performance are as follows: the weight of thrombus location is κ1I, the weight of thrombus size is

[0070] κ2I, thrombus morphology weight is κ3I;

[0071] And according to each weight and external and internal manifestations, treatment methods are presented;

[0072] Or, if there are multiple treatment methods, set the external and internal floating weight influence factors F o and F I ,

[0073]

[0074] Then change the weight,

[0075] The weights of each external manifestation are: The redness weight is λ2O z , the skin temperature weight is λ3O z , the color weight of the skin is

[0076]

[0077] The weights of each intrinsic performance are as follows: The weight of thrombus location is The weight of thrombus size is The weight of thrombus morphology was κ3I;

[0078] And according to the weights and external and internal manifestations, treatment methods are given.

[0079] If after the above adjustments, there are still multiple treatment methods, continue to use the above method to change the weights according to the importance, secondary importance, and secondary importance of the external and internal manifestations until the most appropriate treatment method is obtained, or stop when the number of changes reaches the set number (considered). At this time, even multiple methods can be recommended to the patient.

[0080] The prevention model is a treatment recommendation model established based on the symptoms, treatment methods used and recovery status of patients with lower limb venous thrombosis in the lower limb venous thrombosis data model. Through the prevention model, patients can understand the treatment methods that need to be taken at this stage, and then make a comprehensive choice based on the doctor's diagnosis.

[0081] In the present invention, the prevention model is established based on the data model of lower limb venous thrombosis and covers the treatment methods at different stages of lower limb venous thrombosis.

[0082] The prevention model generates multiple treatment options based on the patient's current condition of lower extremity venous thrombosis, which can be the patient's own feelings and / or the doctor's diagnosis.

[0083] In the prevention model, in a feasible embodiment, in order to enable patients to have a clearer understanding of the treatment methods automatically calculated and recommended by the system, the prevention model ranks multiple treatment methods according to the historical treatment effects of the patients at this stage, and displays them on a display screen or tablet computer as recommendations to patients. Patients can make choices based on their own circumstances or according to the doctor's advice.

[0084] In an achievable embodiment, the percentage of the effect produced by each treatment means is displayed. The percentage is calculated by the formula: (number of cases that produced an effect using the means / total number of cases using the means)*100%.

[0085] In a feasible embodiment, for each treatment measure, the number of times the treatment measure has been selected under the corresponding symptoms is displayed.

[0086] In an achievable embodiment, the symptoms, treatment conditions, and quality of life of new patients are continuously collected to update the lower extremity venous thrombosis database. The historical process, development process, and quality of life model in the lower extremity venous thrombosis data model are updated using a CNN deep learning model. The prevention model is rebuilt based on the updated lower extremity venous thrombosis model, and the treatment methods are updated.

[0087] In a feasible embodiment, after the treatment means are automatically updated by the model, the treatment means and the sequence are manually checked and corrected.

[0088] With the development of medical technology, new treatment methods will bring better treatment effects. In order to enable the system of the present invention to keep pace with the development of technology, the model needs to be updated regularly or irregularly so that it can be displayed to patients in a timely and accurate manner.

[0089] The risk assessment model is used to show patients how blood clots will appear on or in the human body after different treatments are taken, thereby helping patients understand the differences between different treatments and choose different treatments based on their actual circumstances.

[0090] In the risk assessment model, in an achievable embodiment, the risk assessment model displays the external and internal manifestations of the lower limbs in stages based on the treatment method selected by the patient received by the lower limb venous thrombosis data model and the prevention model. The stages are divided into treatment stages or time stages, and the display time of each stage is set to 10 seconds. A total of 5 stages are set, and the patient makes a selection on the display or tablet computer according to the treatment effect to be achieved.

[0091] In the present invention, the lower extremity venous thrombosis data model is based on the symptoms input by the patient or the symptoms selected (if the symptoms are input, keyword matching is performed to correspond to the existing stored symptoms, and multiple symptoms for selection are popped up, allowing the patient to select after entering; if the symptoms entered by the user fail to match the more relevant stored symptoms, the doctor can update them in the background, and update the background data based on the actual situation of the outpatient clinic to make the database more complete), whether there is a family history of thrombosis, whether there is a previous history of thrombosis, whether surgery is performed, and the type of surgery (the type of surgery can be pre-stored in the system, as comprehensive as possible, if it does not exist Doctors can also add it in the background), lifestyle habits (such as smoking, drinking, staying up late, whether sitting for a long time, etc., which are already set in the system and available for selection), etc., to calculate the historical process that the patient experienced when reaching the current state of lower limb venous thrombosis (that is, the change process of the thrombus-producing site from a non-thrombus state to the current thrombus state), the development process of the patient's lower limb venous thrombosis according to the current state (the subsequent change process of the thrombus at this site, and also including the quality of life of patients without treatment) and the subsequent quality of life of patients with lower limb venous thrombosis based on the recommended treatment method (may also include the change process of thrombus under treatment methods). With the help of display equipment, it is presented to the patient in the form of animation, and synchronous voice explanations can be added. Through the combination of vision and hearing, the patient can intuitively feel the current state, the severity of the situation, and the improvement of the correct and necessary treatment (enhance the patient's confidence in treatment), etc.

[0092] It should be noted that, in the present invention, the historical process refers to the process of lower extremity venous thrombosis experienced by the patient before the current condition, including but not limited to the swelling, pain, leg abnormalities, muscle cramps, color changes, skin itching, ulcers, etc. felt by the patient. The development process refers to the process of lower extremity venous thrombosis that the patient will face if he continues his current lifestyle and treatment methods (or no treatment), including possible post-thrombotic syndrome of lower extremity venous thrombosis and pulmonary embolism. The quality of life refers to the physical state, psychological state, etc. of the patient whether the patient adopts treatment methods or not. When establishing the database, the physical state and psychological state are obtained through observation, examination and feedback from patients. The quality of life is an important indicator for judging the condition of patients with lower extremity venous thrombosis.

[0093] It should be noted that the treatment methods referred to in the present invention include physical therapy, drug therapy and surgical treatment. The prevention model also recommends specific treatment methods based on whether the patient has a risk of bleeding. The treatment methods can be single or a combination of multiple methods.

[0094] The beneficial effects of the present invention are as follows: The purpose of the present invention is to provide a visual outpatient device for lower limb venous thrombosis clinics that can be selected and reminded according to the different conditions of patients, so that doctors can have a more appropriate understanding of the doctor's diagnosis and subsequent treatment through the device or system on the basis of traditional medical consultation and explanation. Not only can patients accurately understand the doctor's diagnosis, but also patients can clearly understand the causes of the current condition, and can actively cooperate with the doctor's treatment. There is no misunderstanding between doctors and patients, breaking the barriers of traditional outpatient diagnosis and treatment.

[0095] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A prevention and reminder system for lower extremity venous thrombosis, characterized in that: The prevention reminder system is based on a built-in lower limb venous thrombosis data model. It displays to the patient the historical process of the current lower limb venous thrombosis status, the development process of the patient's lower limb venous thrombosis according to the current status, and the quality of life of the patient with lower limb venous thrombosis based on the recommended treatment method according to the symptoms input or selected by the user, whether there is a family history of thrombosis, whether there is a previous history of thrombosis, whether surgery has been performed and the type of surgery. The prevention reminder system uses a display or a tablet computer to display to the patient.

2. A lower extremity venous thrombosis prevention reminder system according to claim 1, characterized in that: The lower limb venous thrombosis data model is based on a database formed by cases received by the hospital and uses a CNN deep learning model to establish a generation model of the historical process, development progress, and quality of life of lower limb venous thrombosis in historical patients; the input data of the CNN deep learning model are the external manifestations, internal manifestations, and treatment methods of historical patients; the external manifestations include: pain, redness and swelling, skin temperature, and skin color; the internal manifestations include the location of the thrombus, the size of the thrombus, the morphology of the thrombus, and the blood filling degree; the treatment methods include not using treatment methods, the types of treatment methods used, and whether regular examinations are performed.

3. A lower extremity venous thrombosis prevention reminder system according to claim 2, characterized in that: The prevention reminder system is also provided with a prevention model for lower limb venous thrombosis, which is established based on the data model of lower limb venous thrombosis and covers treatment methods for different stages of lower limb venous thrombosis; the prevention model generates multiple treatment methods based on the current condition of the patient's lower limb venous thrombosis, and sorts the multiple treatment methods based on the historical treatment effects of the patients, and displays them on a display screen or tablet computer as a recommendation to the patient; the patient can make a choice based on his or her own situation or according to the doctor's advice.

4. A lower extremity venous thrombosis prevention reminder system according to claim 3, characterized in that: Based on the treatment conditions of new patients, the database is updated, and the CNN deep learning model is used to update the historical process, development process and quality of life model in the lower limb venous thrombosis data model. The prevention model is rebuilt according to the updated lower limb venous thrombosis model, and the treatment methods are updated; then the treatment methods and rankings are manually checked and corrected.

5. A lower extremity venous thrombosis prevention reminder system according to claim 4, characterized in that: The treatment methods include physical therapy, drug therapy and surgical treatment. The prevention model also recommends specific treatment methods based on whether the patient has a risk of bleeding. The treatment methods can be single or a combination of multiple methods.

6. A lower extremity venous thrombosis prevention reminder system according to claim 5, characterized in that: The prevention reminder system is also provided with a risk assessment model, which displays the external and internal manifestations of the lower limbs in stages based on the lower limb venous thrombosis data model and the treatment method selected by the patient received by the prevention model. The stages are divided into treatment stages or time stages, and the display time of each stage is set to 10 seconds. A total of 5 stages are set, and the patient makes a selection on the display or tablet computer according to the treatment effect to be achieved.

7. A lower extremity venous thrombosis prevention reminder system according to claim 6, characterized in that: The historical process refers to the process of lower limb venous thrombosis that the patient has experienced before the current condition; the development process refers to the process of lower limb venous thrombosis that the patient will face if he continues his current lifestyle and treatment methods; the quality of life refers to the physical and psychological state of the patient with or without treatment methods; when establishing the database, the physical and psychological states are obtained through observation, examination and feedback from patients.

8. A lower extremity venous thrombosis prevention reminder system according to claim 7, characterized in that: For the CNN deep learning model, the following formula is used to calculate the initial weight of the external performance: Among them, γ1 represents the pain level at the initial detection, τ1 represents the redness and swelling level at the initial detection, T1 represents the skin temperature at the initial detection, c1 represents the skin color level at the initial detection, λ1, λ2, λ3, λ4 are the parameters of pain, redness and swelling, skin temperature and skin color respectively; The initial weights are corrected, and pain, redness, swelling, skin temperature and skin color are detected multiple times at set intervals, which are recorded as: γ i , τ i , T i 、c i , i = 1, 2, ... n, calculate the difference of the above detection objects. If the average value of the difference is less than the set threshold, the initial weight is used as the weight of the external performance in the CNN deep learning model; as long as the average value of the difference of one of the detection objects is greater than the set threshold, the initial weight needs to be corrected. The calculation formula is: Among them, γ represents the average pain level during detection, τ represents the average redness and swelling level during detection, T represents the average skin temperature during detection, c represents the average skin color level during detection, and Δ is the product of the corresponding parameter and the corresponding difference average value.

9. A lower extremity venous thrombosis prevention reminder system according to claim 8, characterized in that: For the CNN deep learning model, the following formula is used to calculate the weight of the intrinsic representation: Among them, p represents the influencing factor of thrombus position, s represents the influencing factor of thrombus size, η represents the influencing factor of thrombus morphology, ρ represents the blood filling degree, κ1 represents the parameter of the influencing factor of thrombus position, κ2 represents the parameter of the influencing factor of thrombus size, and κ3 represents the parameter of the influencing factor of thrombus morphology.

10. A lower extremity venous thrombosis prevention reminder system according to claim 9, characterized in that: The weights of each external manifestation are: pain weight is λ1O z , redness weight is λ2O z , the skin temperature weight is λ3O z , the skin color weight is λ4O z ; The weights of each intrinsic performance are as follows: the weight of thrombus location is κ1I, the weight of thrombus size is κ2I, and the weight of thrombus morphology is κ3I; And according to each weight and external and internal manifestations, treatment methods are presented; Or, if there are multiple treatment methods, set the external and internal floating weight influence factors F o and F I , Then change the weight, The weights of each external manifestation are: The redness weight is λ2O z , the skin temperature weight is λ3O z , the color weight of the skin is The weights of each intrinsic performance are as follows: The weight of thrombus location is The weight of thrombus size is The weight of thrombus morphology is κ3I; and treatment methods are determined based on the weights and external and internal manifestations.

Images