A lumbar pain prediction system

Abstract

The present invention relates to the field of pain prediction technology, and in particular to a lumbar pain prediction system, the system comprising: a data acquisition and receiving device, a pain prediction device, a data management device, and a data display and feedback device; the data acquisition and receiving device is used to collect multi-dimensional information of system users and receive the best lumbar maintenance plan; the pain prediction device is used to predict the risk level of lumbar pain according to the multi-dimensional information, and then generate a spare lumbar maintenance plan; the data management device is used to store the lumbar maintenance plan, multi-dimensional information, the risk level of lumbar pain, the spare lumbar maintenance plan and the best lumbar maintenance plan; the data display and feedback device is used to generate the best lumbar maintenance plan, and the best lumbar maintenance plan is fed back to the data acquisition and receiving device. The present invention can more accurately predict the risk of lumbar pain in the early stage, guide patients to correct bad habits or seek medical treatment in time, reduce medical costs, and protect the health of patients' lumbar spine.

Description

Technical Field

[0001] The invention relates to the technical field of pain prediction, and in particular to a lumbar pain prediction system. Background Art

[0002] Lumbar pain is a widespread health problem in modern society, affecting the quality of life of a large number of people. Its causes are complex and diverse, including but not limited to lumbar disc herniation, lumbar fracture, lumbar spinal stenosis, lumbar tumors and other diseases. These diseases are often caused by bad habits in daily life, improper working posture or sports injuries, causing long-term pain and inconvenience to patients.

[0003] At present, the diagnosis and prediction of lumbar pain mainly rely on medical imaging technology, such as X-ray examination and lumbar CT scan. Among them, although X-ray examination is widely used, its resolution is low and it is difficult to capture subtle lumbar lesions. In contrast, lumbar CT scan, with its high resolution and multi-layer scanning capabilities, can more accurately locate lesions, show the size, shape and relationship of lesions with surrounding tissues, and become an important means of diagnosing lumbar pain. However, CT scanning is expensive and has certain radiation risks, so it is not suitable as a routine screening method. Therefore, there is no better means of early prediction of the risk of lumbar pain in the prior art. Summary of the invention

[0004] In view of the defects in the prior art, the present invention provides a lumbar pain prediction system.

[0005] In order to achieve the above-mentioned purpose, the present invention provides a lumbar pain prediction system, the system comprising: a data acquisition and receiving device, a pain prediction device, a data management device, and a data display and feedback device; the data acquisition and receiving device is used to collect multi-dimensional information of system users, and can receive the optimal lumbar maintenance plan fed back by the data display and feedback device; the pain prediction device is used to predict the risk level of lumbar pain according to the multi-dimensional information, and then generate a spare lumbar maintenance plan; the data management device is used to store the lumbar maintenance plan, the multi-dimensional information, the lumbar pain risk level, the spare lumbar maintenance plan and the optimal lumbar maintenance plan; medical staff use the data display and feedback device to generate the optimal lumbar maintenance plan, and feed back the optimal lumbar maintenance plan to the data acquisition and receiving device. The present invention can accurately predict the risk of lumbar pain in the early stage, can be used to screen people who may have lumbar pain on a large scale, and guide patients to correct bad habits or seek medical treatment in time, reduce medical costs, and protect patients' lumbar health.

[0006] Optionally, the multi-dimensional information includes collecting individual characteristic data, lifestyle data and work posture data of the patient, the individual characteristic data includes user age, user gender, user weight and user height, the lifestyle data includes user sitting posture data, user standing posture data and user movement data, and the work posture data includes work intensity and work posture.

[0007] Furthermore, the present invention improves the accuracy of early prediction of the user's lumbar pain risk level by collecting multi-dimensional information of system users.

[0008] Optionally, the pain prediction device is used to predict the lumbar pain risk level according to the multi-dimensional information, and then generate a backup lumbar maintenance plan. The pain prediction device specifically performs the following steps:

[0009] Acquire lumbar pain prediction index data according to the multi-dimensional information;

[0010] Inputting the lumbar pain prediction index data into a pain prediction model according to the user's gender to obtain the lumbar pain risk level;

[0011] The lumbar maintenance plan is used to obtain an alternative lumbar maintenance plan according to the lumbar pain risk level.

[0012] Furthermore, the present invention obtains lumbar pain prediction indicator data, and then uses a pain prediction model to predict the lumbar pain risk level, and then generates an alternative lumbar maintenance plan, thereby achieving data-driven risk prediction and plan formulation, making the prediction results more objective and accurate, and providing a basis for obtaining the best lumbar maintenance plan.

[0013] Optionally, the step of obtaining lumbar pain prediction indicator data according to the multi-dimensional information comprises the following steps:

[0014] Preprocessing the multi-dimensional information, and recording the preprocessed multi-dimensional information as standby multi-dimensional information;

[0015] Calculating a BMI value based on the user's weight and the user's height;

[0016] Calculate a waist life pressure index based on the user sitting posture data, the user standing posture data and the user movement data;

[0017] Calculating a waist work pressure index based on the work intensity and the work posture;

[0018] The user age, the BMI value, the lumbar life pressure index and the lumbar work pressure index are used as the lumbar pain prediction indicator data.

[0019] Furthermore, the present invention provides a solid data basis for the subsequent prediction of lumbar pain risk level by acquiring lumbar pain prediction indicator data.

[0020] Optionally, the pain prediction model includes a first pain prediction model and a second pain prediction model, and constructing the pain prediction model includes the following steps:

[0021] Determining the classification criteria of the lumbar pain risk level, and then constructing a first lumbar pain prediction data set and a second lumbar pain prediction data set for males and females respectively;

[0022] Construct a multi-classification support vector machine based on the maximum feature similarity strategy;

[0023] Using the first lumbar pain prediction data set to complete the training and verification of the multi-classification support vector machine to obtain a first pain prediction model;

[0024] The second lumbar pain prediction data set is used to complete the training and verification of the multi-classification support vector machine to obtain a second pain prediction model.

[0025] Furthermore, the present invention takes into account the differences in lumbar health between men and women by constructing pain prediction models for different genders, so that the prediction results are more in line with the actual situation and the accuracy of the prediction is improved.

[0026] Optionally, the construction of a multi-classification support vector machine according to the maximum feature similarity strategy includes the following steps:

[0027] Taking each of the lumbar pain risk levels as the categories that need to be divided by the support vector machine, and combining different categories in pairs to obtain multiple category combinations;

[0028] Calculating the feature difference of each category combination, and merging the category combination with the smallest feature difference into one category;

[0029] After each category merging, different categories are combined in pairs to obtain multiple new category combinations to calculate the feature differences, and then the categories are merged again;

[0030] The categories are continuously merged, and when only one category combination remains, multiple binary classification support vector machines are trained according to the merging order of the categories based on the hierarchical classification method to form the multi-classification support vector machine.

[0031] Furthermore, the multi-classification support vector machine constructed in the present invention has high training efficiency and classification accuracy, is suitable for processing large amounts of data and quickly obtaining prediction results, and is conducive to quickly and accurately predicting the risk level of lumbar pain.

[0032] Optionally, the feature difference satisfies the following relationship:

[0033]

[0034] in, For Category The difference in characteristics of the combination of said categories with category b, For Category The first characteristic difference of the category combination composed of category b, For Category The second characteristic difference of the category combination composed of category b, , is the category in the training sample The number of groups of data occupied, is the number of groups of data occupied by category b in the training sample, is the number of groups of data occupied by the k-th category in the training sample, and the k-th category is based on the category and category b, M is the number of types of lumbar pain prediction index data, is the j-th lumbar pain prediction index data in the i-th group of data in the data occupied by the category combination, and N is the number of data groups in the training sample.

[0035] Furthermore, the first feature difference can reflect the compactness of the data contained in the merged data of different categories. The smaller the first feature difference, the better the unity of the two categories after merging. The second feature difference can reflect the distribution of each category in the overall data space. The smaller the second feature difference, the more difficult it is to separate the categories. Therefore, the feature difference can reflect the overall difference between the two categories, which helps to ensure the accuracy of the classification of the pain prediction model.

[0036] Optionally, the step of using the lumbar maintenance plan to obtain a backup lumbar maintenance plan according to the lumbar pain risk level includes the following steps:

[0037] The pain prediction device selects a plurality of lumbar maintenance plans from the data management device as target lumbar maintenance plans according to the lumbar pain risk level;

[0038] The pain prediction device combines the target lumbar maintenance plan to generate the backup lumbar maintenance plan.

[0039] Furthermore, the present invention generates a backup lumbar maintenance plan by screening out multiple lumbar maintenance plans from a data management device according to the risk level of lumbar pain. This flexible plan generation mechanism can be personalized according to the user's specific circumstances, thereby improving the pertinence and effectiveness of the backup lumbar maintenance plan.

[0040] Optionally, the medical staff generates the optimal lumbar maintenance plan using the data display and feedback device, and feeds back the optimal lumbar maintenance plan to the data acquisition and receiving device, including the following steps:

[0041] The data display and feedback device displays the multi-dimensional information, the lumbar pain risk level and the backup lumbar maintenance plan to the medical staff;

[0042] The medical staff optimizes the backup lumbar maintenance plan according to the multi-dimensional information and the lumbar pain risk level within a specified time threshold to obtain the optimal lumbar maintenance plan;

[0043] The data display and feedback device attaches an optimized label to the optimal lumbar maintenance plan and then feeds back to the data acquisition and receiving device.

[0044] Furthermore, the present invention optimizes the alternative lumbar maintenance plan according to the opinions of professionals to obtain the best lumbar maintenance plan, so that the lumbar maintenance plan has stronger pertinence and effectiveness, and guides patients to correct bad habits or seek medical treatment in time.

[0045] Optionally, the medical staff generates the optimal lumbar maintenance plan using the data display and feedback device, and feeds back the optimal lumbar maintenance plan to the data acquisition and receiving device, further comprising the following steps:

[0046] If the medical staff fails to optimize the backup lumbar maintenance plan within the time threshold, the lumbar maintenance plan is directly used as the optimal lumbar maintenance plan;

[0047] The data display and feedback device attaches a non-optimized label to the optimal lumbar maintenance plan and then feeds back to the data collection and receiving device.

[0048] Furthermore, when the medical staff fails to optimize the backup lumbar maintenance plan within the time threshold, the backup lumbar maintenance plan is directly fed back to the data collection and receiving device as the optimal lumbar maintenance plan, which helps to improve the communication efficiency between doctors and patients and guide patients to correct bad habits or seek medical treatment in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1A schematic diagram of a framework of a lumbar pain prediction system according to an embodiment of the present invention;

[0051] Figure 2 Schematic diagram of the pain prediction model structure of an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments described herein are only for illustration and are not intended to limit the present invention. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that these specific details do not need to be adopted to implement the present invention. In other examples, in order to avoid confusing the present invention, known circuits, software or methods are not specifically described.

[0053] Throughout the specification, references to "one embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment of the present invention. Therefore, the phrases "in one embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily all refer to the same embodiment or example. In addition, particular features, structures, or characteristics may be combined in one or more embodiments or examples in any suitable combination and / or subcombination. In addition, it should be understood by those of ordinary skill in the art that the figures provided herein are for illustrative purposes and that the figures are not necessarily drawn to scale.

[0054] It should be noted in advance that, in an optional embodiment, except for independent explanations, the same symbols or letters appearing in all formulas have the same meanings and values.

[0055] In an alternative embodiment, see Figure 1 The present invention provides a lumbar pain prediction system, which includes: a data acquisition and receiving device 1, a pain prediction device 2, a data management device 3, and a data display and feedback device 4.

[0056] Data transmission is achieved among the data collection and receiving device 1, the pain prediction device 2, the data management device 3, and the data display and feedback device 4 through the Internet of Things. The data collection and receiving device 1 is used to collect multi-dimensional information of system users, and can receive the optimal lumbar maintenance plan fed back by the data display and feedback device 4; the pain prediction device 2 is used to predict the lumbar pain risk level according to the multi-dimensional information, and then generate a backup lumbar maintenance plan; the data management device 3 is used to store the lumbar maintenance plan, the multi-dimensional information, the lumbar pain risk level, the backup lumbar maintenance plan, and the optimal lumbar maintenance plan; medical staff use the data display and feedback device 4 to generate the optimal lumbar maintenance plan, and feed back the optimal lumbar maintenance plan to the data collection and receiving device 1.

[0057] In an optional embodiment, the data collection and receiving device 1 is a mobile phone with a lumbar pain prediction app installed, and the user uploads his own multi-dimensional information through the lumbar pain prediction app, and queries the lumbar pain risk level and the best lumbar maintenance plan. Further, in other optional embodiments, the user can also use the WeChat applet to upload his own multi-dimensional information, and query the lumbar pain risk level and the best lumbar maintenance plan.

[0058] Specifically, in this embodiment, the multi-dimensional information includes collecting the patient's individual characteristic data, life habit data and work posture data. The individual characteristic data includes user age, user gender, user weight and user height. The life habit data includes user sitting posture data, user standing posture data and user movement data. The work posture data includes work intensity and work posture.

[0059] In an optional embodiment, the pain prediction device 2 is specifically a computer, and the pain prediction device 2 receives multi-dimensional information collected by the data collection and receiving device 1 through the Internet of Things. This embodiment obtains lumbar pain prediction index data, and then uses the pain prediction model to predict the lumbar pain risk level, and then generates a backup lumbar maintenance plan, thereby realizing data-driven risk prediction and plan formulation, making the prediction results more objective and accurate, and providing a basis for obtaining the best lumbar maintenance plan. The pain prediction device 2 specifically performs the following steps:

[0060] S1. Obtain lumbar pain prediction indicator data based on the multi-dimensional information.

[0061] Wherein, step S1 specifically includes the following steps:

[0062] S11. Preprocess the multi-dimensional information, and record the preprocessed multi-dimensional information as standby multi-dimensional information.

[0063] Specifically, in this embodiment, the process of preprocessing the multi-dimensional information includes supplementing missing values ​​and eliminating abnormal values.

[0064] S12. Calculate a BMI value based on the user's weight and the user's height.

[0065] Specifically, the calculation of BMI value is a prior art.

[0066] S13. Calculate a waist life pressure index based on the user sitting posture data, the user standing posture data and the user movement data.

[0067] Specifically, in this embodiment, the user's sitting posture data includes the angle between the user's upper body and the horizontal plane when the user is sitting and the daily sitting time, the user's standing posture data includes the angle between the user's upper body and the horizontal plane when the user is standing and the daily standing time, and the user's exercise data includes the exercise type and the daily exercise time. The waist life pressure index satisfies the following relationship:

[0068]

[0069] in, is the waist life pressure index, The length of time you sit or stand each day. is the angle between the user's upper body and the horizontal plane when the user is sitting. The daily standing time, is the angle between the user's upper body and the horizontal plane when the user is standing, The weight of the damage to the waist caused by different types of sports, The duration of daily exercise. The value is based on expert evaluation method, and the daily sitting time or daily standing time does not include the daily exercise time.

[0070] S14. Calculate a waist work pressure index based on the work intensity and the work posture.

[0071] Specifically, in this embodiment, the user's weight and working time are used to reflect the work intensity, and the angle between the user's upper body and the horizontal plane during work is used to reflect the working posture, and the waist work pressure index satisfies the following relationship:

[0072]

[0073] in, is the waist working pressure index, Bear the weight for the user. The daily working hours are It is the angle between the user's upper body and the horizontal plane when working. Daily working time is not included in daily sitting time or daily standing time.

[0074] S15. Use the user age, the BMI value, the lumbar life stress index and the lumbar work stress index as lumbar pain prediction indicator data.

[0075] Specifically, in this embodiment, the lumbar life stress index and the lumbar work stress index are both simple calculations, the purpose of which is to improve the efficiency of predicting the risk level of lumbar pain.

[0076] Furthermore, this embodiment provides a solid data basis for the subsequent prediction of lumbar pain risk level by obtaining lumbar pain prediction indicator data.

[0077] S2. Input the lumbar pain prediction index data into a pain prediction model according to the user's gender to obtain the lumbar pain risk level.

[0078] Specifically, in this embodiment, the pain prediction model includes a first pain prediction model and a second pain prediction model, and constructing the pain prediction model includes the following steps:

[0079] E1. Determine the classification standard of the lumbar pain risk level, and then construct a first lumbar pain prediction data set and a second lumbar pain prediction data set for males and females respectively.

[0080] Specifically, in this embodiment, the classification criteria for lumbar pain risk levels are shown in Table 1:

[0081] Table 1 Classification criteria for lumbar pain risk levels

[0082] Lumbar pain risk level The probability of lumbar pain Level 1 (low risk) [0，0.1) Level 2 (Medium Risk) [0.1，0.3) Level 3 (high risk) [0.3，0.5) Level 4 (extremely high risk) [0.5，1)

[0083] Furthermore, for males, the user age, BMI value, lumbar life pressure index, lumbar work pressure index and whether lumbar pain occurs are collected from the target population. Then, the proportion of people with lumbar pain in the target population with the same user age, BMI value, lumbar life pressure index and lumbar work pressure index is used as the probability of lumbar pain. Finally, the lumbar pain risk level under the user age, BMI value, lumbar life pressure index and lumbar work pressure index is determined according to the classification standard of lumbar pain risk level, and then the first lumbar pain prediction data set is constructed.

[0084] Furthermore, for females, a second lumbar pain prediction data set is obtained in the same manner as the first lumbar pain prediction data set.

[0085] E2. Construct a multi-classification support vector machine based on the maximum feature similarity strategy.

[0086] Wherein, step E2 specifically includes the following steps:

[0087] E21. Taking each of the lumbar pain risk levels as the categories that need to be divided by the support vector machine, and combining different categories in pairs to obtain multiple category combinations.

[0088] Specifically, in this embodiment, the lumbar pain risk level is divided into 4 levels according to the classification standard of the lumbar pain risk level, so 6 category combinations can be obtained.

[0089] E22. Calculate the feature difference of each category combination, and merge the category combination with the smallest feature difference into one category.

[0090] Specifically, in this embodiment, the feature difference satisfies the following relationship:

[0091]

[0092] in, For Category The difference in characteristics between the category combination composed of category a and category b, For Category The first characteristic difference of the category combination with category b, For Category The second characteristic difference of the category combination with category b, , is the category in the training sample The number of groups of data occupied, is the number of groups of data occupied by category b in the training sample, is the number of groups of data occupied by the k-th category in the training sample, and the k-th category is based on the category and category b, M is the number of types of lumbar pain prediction indicator data, is the j-th lumbar pain prediction index data in the i-th group of data in the data occupied by the category combination, and N is the number of data groups in the training sample.

[0093] Furthermore, two categories in a category combination with the smallest feature difference are merged into a new category. This process can refer to the category merging process of the support vector machine based on the hierarchical classification method, and will not be described in detail here.

[0094] The first feature difference can reflect the compactness of the data contained in the merged data of different categories. The smaller the first feature difference, the better the unity of the two categories after merging. The second feature difference can reflect the distribution of each category in the overall data space. The smaller the second feature difference, the more difficult it is to separate the categories. Therefore, the feature difference can reflect the overall difference between the two categories, which helps to ensure the accuracy of the classification of the pain prediction model.

[0095] E23. After each category merging, different categories are combined in pairs to obtain multiple new category combinations to calculate the feature differences, and then the categories are merged again.

[0096] Specifically, in this embodiment, if the number of categories initially obtained is p, after merging a category combination with the smallest feature difference into one category, the number of categories will be reduced by one, that is, p-1. The remaining p-1 categories are used for pairwise combination to obtain multiple new category combinations and calculate the feature difference of each category combination, and the category combination with the smallest feature difference is merged into one category. After the second category merger, the number of categories will be reduced by one again, that is, p-2, and so on.

[0097] E24. Continuously merge categories, and when only one category combination remains, train multiple binary classification support vector machines according to the merged order of the categories based on the hierarchical classification method to form the multi-classification support vector machine.

[0098] Specifically, in this embodiment, the category combination with the smallest feature difference is continuously merged according to the operation of step E23 until only two categories remain, that is, only one category combination remains, and then multiple binary classification support vector machines are trained according to the merging order of the categories based on the hierarchical classification method to form the multi-classification support vector machine.

[0099] Furthermore, based on the hierarchical classification method, multiple binary support vector machines are trained in the order of merging categories to form a multi-classification support vector machine, which actually generates a multi-classification decision tree similar to a binary tree based on the support vector machine. The construction method of the multi-classification support vector machine of this embodiment can automatically determine the structure of the multi-classification support vector machine according to the sample characteristics, thereby improving the training efficiency and classification accuracy.

[0100] E3. Use the first lumbar pain prediction data set to complete the training and verification of the multi-classification support vector machine to obtain a first pain prediction model.

[0101] Specifically, in this embodiment, the first lumbar pain prediction data set is divided into a training set and a validation set in a ratio of 7:3, and then the training set and the validation set are used to complete the training and validation of the multi-classification support vector machine to obtain the first pain prediction model.

[0102] Further, the final first pain prediction model can be found in Figure 2 , Figure 2 Each SVM in represents a binary support vector machine. The first pain prediction model can output four categories, namely, category 1, category 2, category 3 and category 4, which correspond to four lumbar pain risk levels.

[0103] E4. Use the second lumbar pain prediction data set to complete the training and verification of the multi-classification support vector machine to obtain a second pain prediction model.

[0104] Specifically, in this embodiment, the implementation process of this step can refer to step E3, which will not be described in detail here.

[0105] Furthermore, the second pain prediction model finally obtained is the same as the first pain prediction model, specifically as follows Figure 2 shown.

[0106] S3. Use the lumbar maintenance plan to obtain an alternative lumbar maintenance plan based on the lumbar pain risk level.

[0107] Wherein, step S3 specifically includes the following steps:

[0108] S31. The pain prediction device selects a plurality of lumbar maintenance plans from the data management device as target lumbar maintenance plans according to the lumbar pain risk level.

[0109] Specifically, in this embodiment, the data management device 3 stores lumbar maintenance plans suitable for different lumbar pain risk levels according to different user ages, BMI values, lumbar life stress indexes and lumbar work stress indexes. These lumbar maintenance plans can be collected through the Internet or set by experts.

[0110] More specifically, for the user's age, BMI value, lumbar life stress index and lumbar work stress index determined under a certain lumbar pain risk level, the pain prediction device 2 respectively selects from the data management device 3 a total of four lumbar maintenance plans that are suitable for the user's age, BMI value, lumbar life stress index and lumbar work stress index under the lumbar pain risk level as target lumbar maintenance plans.

[0111] S32. The pain prediction device combines the target lumbar maintenance plan to generate the backup lumbar maintenance plan.

[0112] This embodiment generates a backup lumbar maintenance plan by screening out multiple lumbar maintenance plans from the data management device according to the lumbar pain risk level. This flexible plan generation mechanism can be personalized according to the user's specific situation, thereby improving the pertinence and effectiveness of the backup lumbar maintenance plan.

[0113] In an optional embodiment, the data display and feedback device 4 is specifically a computer. The medical staff uses the data display and feedback device 4 to generate the optimal lumbar maintenance plan, and feeds back the optimal lumbar maintenance plan to the data acquisition and receiving device 1, including the following steps:

[0114] F1. The data display and feedback device displays the multi-dimensional information, the lumbar pain risk level and the alternative lumbar maintenance plan to medical staff.

[0115] F2. Medical staff optimizes the backup lumbar maintenance plan according to the multi-dimensional information and the lumbar pain risk level within a specified time threshold to obtain the optimal lumbar maintenance plan.

[0116] Specifically, in this embodiment, the specified time threshold is 1 hour. After checking the multi-dimensional information and lumbar pain risk level of the user, the medical staff modifies and supplements the backup lumbar maintenance plan based on their own experience within 1 hour to obtain the best lumbar maintenance plan.

[0117] Furthermore, this embodiment optimizes the alternative lumbar maintenance plan according to the opinions of professionals to obtain the best lumbar maintenance plan, so that the lumbar maintenance plan has stronger pertinence and effectiveness, guiding patients to correct bad habits or seek medical treatment in time, thereby reducing medical costs.

[0118] F3. The data display and feedback device attaches an optimized label to the optimal lumbar maintenance plan and then feeds back to the data collection and receiving device.

[0119] Specifically, in this embodiment, attaching an optimized label to the optimal lumbar maintenance plan can enable users to understand that the optimal lumbar maintenance plan is certified by professionals, thereby improving user satisfaction and facilitating data management.

[0120] F4. If the medical staff fails to optimize the backup lumbar maintenance plan within the time threshold, the backup lumbar maintenance plan will be directly used as the optimal lumbar maintenance plan.

[0121] Specifically, in this embodiment, if the medical staff does not modify or supplement the backup lumbar maintenance plan within 1 hour, the data display and feedback device 4 will directly use the backup lumbar maintenance plan as the optimal lumbar maintenance plan.

[0122] F5. The data display and feedback device attaches a non-optimized label to the optimal lumbar maintenance plan and then feeds back to the data acquisition and receiving device.

[0123] Specifically, in this embodiment, when the medical staff fails to optimize the backup lumbar maintenance plan within the time threshold, the backup lumbar maintenance plan is directly fed back to the data collection and receiving device as the optimal lumbar maintenance plan, which helps to improve the communication efficiency between doctors and patients.

[0124] Furthermore, attaching a non-optimized label to the optimal lumbar maintenance plan can enable users to understand that the optimal lumbar maintenance plan has not been certified by professionals, prompting users to refer to the optimal lumbar maintenance plan appropriately to correct bad habits or seek medical treatment, and it is also beneficial for data management.

[0125] It should be noted that, in some cases, the actions described in the specification can be performed in a different order and still achieve the desired results. In this embodiment, the order of steps given is only to make the embodiment appear clearer and easier to explain, rather than to limit it.

[0126] In summary, the present invention obtains lumbar pain prediction index data by collecting multi-dimensional information of users, and then uses the pain prediction model to make early predictions on the risk of lumbar pain, and can formulate the best lumbar maintenance plan based on the prediction results, which helps to screen out users who may have lumbar pain on a large scale, guide patients to correct bad habits or seek medical treatment in time, and protect people's lumbar health. In addition, users can understand the possibility of their own lumbar pain by inputting their own multi-dimensional information, which is also conducive to self-monitoring, correcting bad habits in time, and reducing medical costs.

[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the specification of the present invention.

Claims

1. A lumbar pain prediction system, characterized in that: include: Data collection and receiving equipment, pain prediction equipment, data management equipment, data display and feedback equipment; The data collection and receiving device is used to collect multi-dimensional information of system users, and can receive the optimal lumbar maintenance plan fed back by the data display and feedback device; The multi-dimensional information includes collecting individual characteristic data, life habit data and working posture data of the patient, wherein the individual characteristic data includes user age, user gender, user weight and user height, the life habit data includes user sitting posture data, user standing posture data and user movement data, and the working posture data includes work intensity and working posture; Preprocessing the multi-dimensional information, and recording the preprocessed multi-dimensional information as standby multi-dimensional information; Calculating a BMI value based on the user's weight and the user's height; Calculate a waist life pressure index based on the user sitting posture data, the user standing posture data and the user movement data; Calculating a waist work pressure index based on the work intensity and the work posture; The user's age, the BMI value, the waist life pressure index and the waist work pressure index are used as lumbar pain prediction indicator data; Inputting the lumbar pain prediction index data into a pain prediction model according to the user's gender to obtain a lumbar pain risk level; obtaining an alternate lumbar maintenance plan using a lumbar maintenance plan according to the lumbar pain risk level; The data management device is used to store the lumbar maintenance plan, the multi-dimensional information, the lumbar pain risk level, the backup lumbar maintenance plan and the optimal lumbar maintenance plan; The medical staff generates the optimal lumbar maintenance plan using the data display and feedback device, and feeds back the optimal lumbar maintenance plan to the data acquisition and receiving device; The pain prediction model includes a first pain prediction model and a second pain prediction model, and constructing the pain prediction model includes the following steps: Determining the classification criteria of the lumbar pain risk level, and then constructing a first lumbar pain prediction data set and a second lumbar pain prediction data set for males and females respectively; Taking each of the lumbar pain risk levels as the categories that need to be divided by the support vector machine, and combining different categories in pairs to obtain multiple category combinations; Calculating the feature difference of each category combination, and merging the category combination with the smallest feature difference into one category; After each category merging, different categories are combined in pairs to obtain multiple new category combinations to calculate the feature differences, and then the categories are merged again; The feature differences satisfy the following relationship: ， in, For Category The difference in characteristics of the combination of said categories with category b, For Category The first characteristic difference of the category combination composed of category b, For Category The second characteristic difference of the category combination composed of category b, , is the category in the training sample The number of groups of data occupied, is the number of groups of data occupied by category b in the training sample, is the number of groups of data occupied by the k-th category in the training sample, and the k-th category is based on the category and category b, M is the number of types of lumbar pain prediction index data, is the j-th lumbar pain prediction index data in the i-th group of data in the data occupied by the category combination, and N is the number of groups of data in the training sample; Continuously merge categories, and when only one category combination remains, train multiple binary support vector machines to form a multi-classification support vector machine based on the hierarchical classification method according to the order of category merging; Using the first lumbar pain prediction data set to complete the training and verification of the multi-classification support vector machine to obtain a first pain prediction model; The second lumbar pain prediction data set is used to complete the training and verification of the multi-classification support vector machine to obtain a second pain prediction model.

2. A lumbar pain prediction system according to claim 1, characterized in that: The method of using the lumbar maintenance plan to obtain a backup lumbar maintenance plan according to the lumbar pain risk level comprises the following steps: The pain prediction device selects a plurality of lumbar maintenance plans from the data management device as target lumbar maintenance plans according to the lumbar pain risk level; The pain prediction device combines the target lumbar maintenance plan to generate the backup lumbar maintenance plan.

3. A lumbar pain prediction system according to claim 1, characterized in that: The medical staff generates the optimal lumbar maintenance plan using the data display and feedback device, and feeds back the optimal lumbar maintenance plan to the data acquisition and receiving device, which includes the following steps: The data display and feedback device displays the multi-dimensional information, the lumbar pain risk level and the backup lumbar maintenance plan to the medical staff; The medical staff optimizes the backup lumbar maintenance plan according to the multi-dimensional information and the lumbar pain risk level within a specified time threshold to obtain the optimal lumbar maintenance plan; The data display and feedback device attaches an optimized label to the optimal lumbar maintenance plan and then feeds back to the data acquisition and receiving device.

4. A lumbar pain prediction system according to claim 3, characterized in that: The medical staff generates the optimal lumbar maintenance plan using the data display and feedback device, and feeds the optimal lumbar maintenance plan back to the data acquisition and receiving device, which also includes the following steps: If the medical staff fails to optimize the backup lumbar maintenance plan within the time threshold, the backup lumbar maintenance plan is directly used as the optimal lumbar maintenance plan; The data display and feedback device attaches an unoptimized label to the optimal lumbar maintenance plan and then feeds back to the data acquisition and receiving device.

Images