Ankylosing spondylitis comprehensive management system

Abstract

The present application relates to the technical field of data processing, in particular to a comprehensive management system for ankylosing spondylitis. The monitoring state values generated by a plurality of continuous historical monitoring time periods corresponding to the current monitoring time are fused, a numerical sequence composed of the plurality of monitoring state values is traversed, and a state value distribution hash table is constructed. According to the state value distribution hash table, a sub-determination value corresponding to each monitoring state value is generated. An abnormality determination value corresponding to the current monitoring time is generated, which avoids errors in the relevant information obtained by a single monitoring time, and thus avoids errors in the evaluation index obtained. The present application overcomes the problem in the prior art that although the questionnaire can be used to collect various related symptom information of the AS patient in the current period, the relevant information deviation is generated due to the influence of the subjective feeling of the user and other unexpected situations, and thus a large error is generated in the evaluation index, which cannot be accurately warned.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a comprehensive management system for ankylosing spondylitis. Background Technology

[0002] Ankylosing spondylitis (AS) is a highly disabling, lifelong disease that can lead to disability due to spinal osteophyte formation and hip joint destruction. Previous data shows that up to 40% of AS patients eventually experience complete spinal fusion, even resulting in extreme physical distortion. Because AS primarily affects young men, it has a greater impact on young men who engage in high-intensity physical activities, such as military personnel or athletes. The spinal pain and stiffness caused by AS typically directly affect training and daily life, increasing the risk of training injuries and leading to longer treatment periods and higher retirement rates, making it a significant factor contributing to attrition.

[0003] In existing technologies, questionnaires can be used to collect various relevant symptom information from AS patients at the current time, such as the degree of pain in relevant areas, the degree of fatigue, and the duration of morning stiffness. Appropriate algorithms are then used to process this information to obtain an evaluation index of AS severity, which can then be used to provide patients with corresponding risk warnings. However, because the information collection process can be influenced by the user's subjective feelings and other unforeseen circumstances, the collected information can easily become biased, leading to significant errors in the obtained evaluation index and making accurate warnings impossible. Summary of the Invention

[0004] To address the aforementioned technical problems, the technical solution adopted by this invention is as follows:

[0005] According to one aspect of the present invention, a comprehensive management system for ankylosing spondylitis is provided, the system comprising: a determination module;

[0006] The judgment module is used to fuse the monitoring status values ​​generated by multiple consecutive historical monitoring periods corresponding to the current monitoring time to generate the anomaly judgment value corresponding to the current monitoring time. The monitoring status values ​​include a first type of status value that indicates that the target event is in a normal state; and a second type of status value and a third type of status value that indicate that the target event is abnormal. The degree of abnormality of the target event corresponding to the second type of status value is greater than the degree of abnormality of the target event corresponding to the third type of status value.

[0007] The fusion process includes:

[0008] Traverse the numerical sequence composed of multiple said monitoring status values to construct a status value distribution hash table; the keys of the status value distribution hash table are the first type of status value or the second type of status value or the third type of status value, and each type of status value corresponds to a status value position sequence. Among them, the status value position sequence Key corresponding to the i-th type of status value i =(wz i 1, wz i 2, …, wz i m, …, wz i f(i)), wz i m is the position serial number of the m-th i-th type of status value in the numerical sequence; f(i) is the total number of the i-th type of status value in the numerical sequence; m = 1, 2, …, f(i); i = 1 or 2 or 3;

[0009] Generate a sub-judgment value corresponding to each monitoring status value according to the status value distribution hash table;

[0010] Among them, the sub-judgment value P corresponding to the i-th monitoring status value i Satisfies the following conditions:

[0011]

[0012] Among them, S i Is the number of times the i-th monitoring status value appears in the numerical sequence, S sum Is the total number of monitoring status values in the numerical sequence, μ w Is the mean value of the position distribution of the i-th monitoring status value in the numerical sequence, Is the density of the position distribution of the i-th monitoring status value in the numerical sequence, P i 1 Is the abnormal score corresponding to the i-th monitoring status value;

[0013] Generate an abnormal judgment value P corresponding to the current monitoring moment according to the sub-judgment values P1, P2, and P3 corresponding to the first type of status value, the second type of status value, and the third type of status value respectively; P satisfies the following condition: P = P3 + P2 - P1.

[0014] Furthermore, the fusion processing further includes:

[0015] If P > Y1, generate an abnormal reminder message; Y1 is the abnormal threshold.

[0016] Furthermore, the fusion processing further includes:

[0017] If P < Y1, do not generate an abnormal reminder message; Y1 is the abnormal threshold.

[0018] Furthermore, it further includes: an information acquisition module;

[0019] The information acquisition module is used to acquire the feature observation sequence corresponding to each monitoring period. The feature observation sequence includes attribute values ​​of at least one observable dimension corresponding to the target event.

[0020] Furthermore, it also includes a target hidden Markov model, which is communicatively connected to the information acquisition module and the decision module, respectively.

[0021] The target hidden Markov model is used to generate the monitoring status value corresponding to each historical monitoring period based on the characteristic observation sequence of multiple consecutive historical monitoring periods corresponding to the current monitoring time.

[0022] Furthermore, the target Hidden Markov Model is obtained as follows:

[0023] Collect historical information on target events across multiple consecutive historical monitoring periods; the historical information includes the characteristic observation sequences and monitoring status values ​​corresponding to the target events.

[0024] The initial Hidden Markov Model is trained using historical information about the target event to generate the target Hidden Markov Model.

[0025] Furthermore, it also includes a model update module, which is used for...

[0026] Obtain user feedback on the accuracy of each anomaly prediction generated based on the anomaly determination value; the feedback includes whether the prediction is accurate, the prediction is too high, or the prediction is too low.

[0027] when At this time, the user feature update training set is generated by using historical information of target events in multiple consecutive historical monitoring periods generated by the current user within the current update cycle; G1 is the number of feedback information obtained from the last update to the current period that is accurate for prediction information; G is the total number of feedback information obtained from the last update to the current period; Y2 is the model update threshold.

[0028] The monitoring state value Q in the user feature update training set satisfies the following condition: Q = Q1 × K;

[0029] Where Q1 is the original anomaly prediction information corresponding to Q; K is the adjustment coefficient of the user's feedback information to Q1. When the feedback information is that the prediction information is accurate, K = 1; when the feedback information is that the prediction information is too high, K = 0.8; when the feedback information is that the prediction information is too low, K = 1.2.

[0030] The target Hidden Markov Model is updated and trained using the updated training set with user features.

[0031] Further observable dimensions include the degree of pain and weakness in the neck, back, or hips.

[0032] The present invention has at least the following beneficial effects:

[0033] This invention acquires multiple monitoring status values ​​over a continuous period to form multiple status levels corresponding to a target event within a certain time period. Simultaneously, by analyzing the distribution location, density, and proportion of monitoring status values ​​for three different abnormal states, the invention calculates the proportion and approximate distribution of each type of abnormal state in the entire numerical sequence, thereby reflecting the changing trend of the target event within the monitoring period. Furthermore, by considering the calculation of status values ​​from multiple monitoring periods to generate the final anomaly determination value, the invention mitigates the impact of errors in the currently generated status values ​​on the final anomaly determination value when relying solely on the currently generated status value. This approach minimizes the risk of significant errors in the AS severity evaluation index due to errors in information acquired from a single monitoring time period, thus improving the accuracy of early warnings. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A structural block diagram of a comprehensive management system for ankylosing spondylitis provided in an embodiment of the present invention;

[0036] Figure 2 A schematic flowchart of a fusion process provided in an embodiment of this application is shown;

[0037] Figure 3 This document illustrates the implementation environment architecture of a comprehensive management system for ankylosing spondylitis provided in an embodiment of this application. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] As one possible embodiment of the present invention, such as Figure 1As shown, a comprehensive management system for ankylosing spondylitis is provided. The system includes: an information acquisition module, a target hidden Markov model, and a determination module. The target hidden Markov model is communicatively connected to the information acquisition module and the determination module, respectively.

[0040] The information acquisition module is used to acquire the feature observation sequence corresponding to each monitoring period. The feature observation sequence includes attribute values ​​of at least one observable dimension corresponding to the target event. In this embodiment, the target event can be the patient's current AS severity. The observable dimension can include multiple symptom manifestations reflecting the severity of AS, such as the degree of pain in the neck, back, or buttocks, the degree of fatigue, and the duration of morning stiffness. The specific dimension settings can be adaptively adjusted according to different populations. In the prior art, there are applications that collect the above information in the form of questionnaires, and the acquisition of feature observation sequences can be achieved by calling other corresponding APIs. In this embodiment, the feature observation sequence can generally be a fixed-length vector, such as (A1, A2, A3, A4, A5, A6), where each element represents the evaluation value of the corresponding symptom.

[0041] The target hidden Markov model is used to generate monitoring status values ​​for each historical monitoring period based on the characteristic observation sequences of multiple consecutive historical monitoring periods corresponding to the current monitoring time. The monitoring status values ​​include a first type of status value representing that the target event is in a normal state; and a second type of status value and a third type of status value representing that the target event is abnormal. The degree of abnormality of the target event corresponding to the second type of status value is greater than the degree of abnormality of the target event corresponding to the third type of status value.

[0042] Specifically, in this embodiment, the monitored state value is the tetanic index, which is used to evaluate the severity of AS. The value ranges from [0,10], and the larger the value, the more severe the AS.

[0043] In this embodiment, the monitored status values ​​can be divided into three types according to the value of the tetany index. The specific classification criteria are as follows:

[0044] Firstly, if the range of the rigidity index is [0, 2.1], then the monitoring status value is the first type of status value; in this embodiment, the first type of status value can indicate that the patient is in a healthy state or that the condition has been relieved or that the disease activity is low.

[0045] Secondly, if the range of the rigidity index is (2.1, 3.5], then the monitoring status value is the second type of status value; in this embodiment, the second type of status value can indicate that the patient is in an unhealthy state and has a high degree of disease activity.

[0046] Third, if the tetanic index ranges from (3.5, 10), the monitoring state value is a third-type state value. In this embodiment, the third-type state value can indicate that the patient is in a severe unhealthy state with extremely high disease activity. In this embodiment, the monitoring state value output by the target hidden Markov model can be the corresponding tetanic index. Then, when forming a numerical sequence composed of multiple monitoring state values, each monitoring state value can be converted into a first-type state value, a second-type state value, or a third-type state value according to the above-mentioned classification criteria.

[0047] In this embodiment, the target Hidden Markov Model is a model that predicts the hidden state sequence (i.e., the tessellation index sequence) corresponding to each of the current multiple feature observation sequences. Specifically, the Viterbi algorithm can be used to find the most likely hidden state sequence among multiple feature observation sequences.

[0048] Since the evaluation indicators of AS severity required in this invention are not directly observable, but are indirectly reflected through a series of observable symptoms and physiological indicators, HMM (Hidden Markov Model) introduces hidden states to model these indirect health states, enabling the model to capture changes in the patient's health status.

[0049] Specifically, the target Hidden Markov Model in this embodiment can be obtained as follows:

[0050] Historical information on target events is collected across multiple consecutive historical monitoring periods. This historical information includes the characteristic observation sequences and monitoring status values ​​corresponding to the target events.

[0051] Specifically, historical information can be constructed according to Table 1 below:

[0052] Table 1

[0053]

[0054] The initial Hidden Markov Model is trained using historical information about the target event to generate the target Hidden Markov Model.

[0055] After collecting a large amount of historical information, the parameters of the HMM model can be estimated using the Baum-Welch algorithm (also known as the forward-backward algorithm, which is an unsupervised learning algorithm) or the Viterbi training algorithm (supervised learning algorithm). That is, the parameters of the initial state distribution π, the state transition matrix A, and the emission matrix B are determined so that the model can better fit the observed data.

[0056] In addition, in order to enable the HMM to generate more accurate prediction results based on the user's characteristics when targeting different users, the system in this embodiment also includes a model update module. The model update module is used to obtain feedback information from the user on the accuracy of each abnormal prediction information generated based on the abnormality judgment value; the feedback information includes the prediction information being accurate, the prediction information being too high, and the prediction information being too low.

[0057] when When using the historical information of target events generated by the current user in multiple consecutive historical monitoring periods within the current update cycle, a user feature update training set is generated; G1 is the number of feedback information obtained from the last update to the current period that is accurate in predicting information; G is the total number of feedback information obtained from the last update to the current period; Y2 is the model update threshold; in order to obtain more training samples, the value of G should be greater than 100.

[0058] The monitoring state value Q in the user feature update training set satisfies the following condition: Q = Q1 × K;

[0059] Where Q1 is the original anomaly prediction information corresponding to Q; K is the adjustment coefficient of the user's feedback information to Q1. When the feedback information is that the prediction information is accurate, K = 1; when the feedback information is that the prediction information is too high, K = 0.8; when the feedback information is that the prediction information is too low, K = 1.2.

[0060] The target Hidden Markov Model is updated and trained using the updated training set with user features.

[0061] Different users typically exhibit different physical states. Therefore, to make the predictions of the Hidden Model (HMM) more closely reflect the user's actual situation, the HMM can be fine-tuned using the user's own data. At the same time, by obtaining user feedback, the model's predictions can be fine-tuned to make the hidden state values ​​in the training set updated with user features more closely reflect the user's actual performance, thereby improving the prediction accuracy of the model after retraining.

[0062] The judgment module is used to fuse the monitoring status values ​​generated by multiple consecutive historical monitoring periods corresponding to the current monitoring time to generate the anomaly judgment value corresponding to the current monitoring time.

[0063] like Figure 2 As shown, the fusion process includes:

[0064] S100: Traverse the numerical sequence consisting of multiple monitored state values ​​and construct a state value distribution hash table. The key of the state value distribution hash table is either a first-type state value, a second-type state value, or a third-type state value. Each type of state value corresponds to a state value position sequence, where the key of the state value position sequence corresponding to the i-th type of state value is...i =(wz i 1. wz i 2、…、wz i m、…、wz i f(i)), wz i m is the position number of the m-th i-th type state value in the numerical sequence; f(i) is the total number of the i-th type state values ​​in the numerical sequence; m = 1, 2, ..., f(i); i = 1, 2, or 3; that is, the types of monitored state values ​​in this embodiment are the first type state value, the second type state value, or the third type state value.

[0065] S200: Generate a sub-judgment value corresponding to each monitoring state value based on the state value distribution hash table;

[0066] Among them, the sub-judgment value P corresponding to the i-th monitoring state value i The following conditions must be met:

[0067]

[0068] Among them, S i S represents the number of times the i-th monitoring state value appears in the numerical sequence. sum μ represents the total number of monitored state values ​​in the numerical sequence. w Let be the mean value of the positional distribution of the i-th monitoring state value in the numerical sequence. P represents the density of the positional distribution of the i-th monitoring state value in the numerical sequence. i 1 is the abnormal score corresponding to the i-th monitoring status value.

[0069] In this embodiment, by calculating the mean of the positional distribution of each monitoring state value in the numerical sequence, the approximate positional distribution of that monitoring state value in the numerical sequence can be determined. This also determines whether the occurrence time of that monitoring state is close to the current time. In this embodiment, the larger the rank of the monitoring state value in the numerical sequence, the closer it is to the current time. Therefore, a larger mean of positional distribution indicates that it is closer to the current monitoring time. The closer it is to the current monitoring time, the closer the state is to the current true state, and thus it should be given more weight. Simultaneously, the standard deviation σ of the positional distribution is calculated... w The density of the monitoring status value distribution is used to determine the degree of density. Generally, a higher density (i.e., a lower standard deviation) indicates that the health status will appear more densely around the location distribution mean. Therefore, the ratio of the location distribution mean to the location distribution density reflects the reliability of the location distribution mean for each status. In other words, the location distribution density can be used to further adjust the size of the location distribution mean, making the calculation results more accurate.

[0070] S300: Generate an abnormal determination value P corresponding to the current monitoring moment according to the sub-determination values P1, P2, and P3 corresponding to the first type of status value, the second type of status value, and the third type of status value respectively. P satisfies the following condition: P = P3 + P2 - P1.

[0071] Furthermore, the fusion processing further includes:

[0072] S400: If P > Y1, generate an abnormal reminder message. If P < Y1, do not generate an abnormal reminder message. Y1 is an abnormal threshold. The specific value of Y1 can be determined according to the actual scenario.

[0073] In the present invention, by obtaining multiple monitoring status values obtained in a continuous period of time, multiple status degrees corresponding to the target event within a certain period of time are formed. At the same time, by considering the distribution position, distribution density, and quantity ratio of the monitoring status values of three different abnormal states, the proportion and approximate distribution position of various types of abnormal states in the entire numerical sequence can be calculated, thereby reflecting the change trend of the target event during the monitoring period. And by considering calculating the status values of different multiple monitoring periods to generate the final abnormal determination value, the influence amplitude of the error of the currently generated status value on the error of the final abnormal determination value can be reduced when only the currently generated status value is directly used as the abnormal determination value. Furthermore, it is possible to avoid as much as possible the problem that a large error occurs in the evaluation index of the AS severity due to the error of the relevant information obtained at a single monitoring time, so as to improve the accuracy of early warning.

[0074] As another possible embodiment of the present invention, the technical solutions in the above embodiments can also be applied to the detection of the remaining service life of related hardware devices in a computer, such as in the monitoring scenario of the remaining service life of a hard disk. Specifically, the remaining service life of the hard disk (i.e., the time to the next failure) belongs to a value that cannot be directly observed. However, the values of multiple characteristic dimensions used to indicate the current operating state of the hard disk can be directly observed and obtained, such as the hard disk operating temperature, raw read error rate, start / stop count, command timeout count, etc. Similarly, in actual data collection, due to the failure of the corresponding sensor, some or even all of the collected characteristic data may be in error. In the prior art, the method of predicting the remaining service life of the hard disk based on the currently collected characteristics will result in false alarms.

[0075] Based on this, the methods of S100 to S400 used by the determination module in the present invention can be used in the monitoring scenario of the remaining service life of the hard disk to reduce the probability of false alarms.

[0076] Specifically, in this embodiment, the monitored status value is the operating status index (equivalent to the convulsion index in the above embodiment). This operating status index is the reciprocal of the actual number of days corresponding to the remaining lifespan of the hard drive. In this embodiment, the value range of the actual number of days corresponding to the remaining lifespan of the hard drive is [1, 30]. The monitoring status value can be divided into three types using the operating status index. The specific classification criteria are as follows:

[0077] Firstly, if the range of the operating status index is (0, 1 / 30], then the monitoring status value is the first type of status value; in this embodiment, the first type of status value can indicate that the hard drive is in a healthy state and will not fail in the near future.

[0078] Secondly, if the operating status index ranges from (1 / 30 to 1 / 10), then the monitoring status value is the second type of status value. In this embodiment, the second type of status value can indicate that the hard drive is in an unhealthy state and is prone to failure in the near future.

[0079] Third, if the range of the rigidity index is (1 / 10, 1], then the monitoring status value is the third type of status value. In this embodiment, the third type of status value can indicate that the hard drive is in a state of extreme vulnerability to damage.

[0080] In addition, in order to make the range of values ​​of the running status index in this embodiment closer to the range of values ​​of the rigidity index in the above embodiment for easier subsequent calculation, the range of values ​​of the running status index can also be normalized so that its range becomes (0, 10).

[0081] In this embodiment, the feature dimensions in the corresponding feature observation sequence may include hard disk operating temperature, raw read error rate, start / stop count, command timeout count, and other feature values ​​that reflect the hard disk operating status.

[0082] Therefore, after processing by methods S100 to S400, the continuous operational health of the hard drive within a certain time period can be determined by acquiring multiple monitoring status values ​​obtained over a continuous period. Simultaneously, by analyzing the distribution location, density, and proportion of monitoring status values ​​for three different abnormal states, the percentage and approximate distribution of each type of abnormal state within the entire numerical sequence can be calculated. This reflects the changing trend of the hard drive's health status within the monitoring period. Furthermore, this avoids errors in observational information obtained from a single monitoring time, which could lead to significant errors in the resulting hard drive health evaluation indicators, thus improving the accuracy of early warnings.

[0083] Furthermore, although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0084] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, mobile terminal, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0085] In an exemplary embodiment of this disclosure, an electronic device capable of implementing the above-described method is also provided.

[0086] Those skilled in the art will understand that various aspects of the present invention can be implemented as systems, methods, or program products. Therefore, various aspects of the present invention can be specifically implemented in the following forms: entirely hardware implementations, entirely software implementations (including firmware, microcode, etc.), or implementations combining hardware and software aspects, collectively referred to herein as “circuits,” “modules,” or “systems.”

[0087] An electronic device according to this embodiment of the invention. The electronic device is merely an example and should not be construed as limiting the functionality or scope of the embodiments of the invention.

[0088] Electronic devices are manifested in the form of general-purpose computing devices. Components of an electronic device may include, but are not limited to: at least one processor, at least one memory, and buses connecting different system components (including memory and processor).

[0089] The memory stores program code that can be executed by a processor, causing the processor to perform the steps described in the "Exemplary Methods" section above, according to various exemplary embodiments of the present invention.

[0090] The storage may include readable media in the form of volatile storage, such as random access memory (RAM) and / or cache memory, and may further include read-only memory (ROM).

[0091] The storage may also include programs / utilities having a set (at least one) of program modules, including but not limited to: an operating system, one or more applications, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0092] A bus can represent one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus that uses any of the various bus architectures.

[0093] The electronic device can also communicate with one or more external devices (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable a user to interact with the electronic device, and / or any device that enables the electronic device to communicate with one or more other computing devices (e.g., routers, modems, etc.). This communication can be performed via input / output (I / O) interfaces. Furthermore, the electronic device can communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter. The network adapter communicates with other modules of the electronic device via a bus. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with the electronic device, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0094] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0095] refer to Figure 3 , Figure 3 A schematic diagram of the structure of a computer system suitable for implementing an electronic device or server according to embodiments of this application is shown.

[0096] like Figure 3As shown, a computer system includes a central processing unit (CPU), which performs various appropriate actions and processes based on programs stored in read-only memory (ROM) or loaded from memory into random access memory (RAM). RAM also stores various programs and data required for the system's operating instructions. The CPU, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0097] The following components are connected to the I / O interface: input sections including keyboards, mice, etc.; output sections including cathode ray tubes (CRTs), liquid crystal displays (LCDs), and speakers; storage sections including hard disks; and communication sections including network interface cards such as LAN cards and modems. The communication section performs communication processing via a network such as the Internet. Drives are also connected to the I / O interface as needed. Removable media, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on the drive as needed so that computer programs read from them can be installed into the storage section as required.

[0098] Specifically, according to embodiments of this application, the flowchart above refers to... Figure 2 The described process can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowchart. In such an embodiment, the computer program contains program code for performing the methods shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the functions defined in the system of this application.

[0099] In exemplary embodiments of this disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the methods described above is stored. In some possible embodiments, various aspects of the present invention may also be implemented as a program product comprising program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps of the various exemplary embodiments of the present invention described in the "Exemplary Methods" section above.

[0100] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0101] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0102] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0103] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0104] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0105] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0106] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An ankylosing spondylitis comprehensive management system, characterized by, The system comprises a determination module; The determination module is configured to fuse monitoring state values generated in a plurality of continuous historical monitoring periods corresponding to a current monitoring time, to generate an abnormality determination value corresponding to the current monitoring time; the monitoring state values comprise first type state values respectively indicating that a target event is in a normal state, and second type state values and third type state values indicating that the target event is abnormal, the second type state values corresponding to a greater degree of abnormality of the target event than the third type state values; The fusion processing comprises: traversing the numerical sequence composed of the plurality of monitoring state values, a state value distribution hash table is constructed; the key of the state value distribution hash table is a first type state value or a second type state value or a third type state value, each type state value corresponds to a state value position sequence, wherein the i-th type state value corresponds to a state value position sequence Key i =(wz i 1、wz i 2、…、wz i m、…、wz i f(i)),wz i m is the position number of the m-th i-th type state value in the numerical sequence; f(i) is the total number of the i-th type state value in the numerical sequence; m=1, 2, …, f(i); i=1 or 2 or 3; generating a sub-determination value corresponding to each type of monitoring state value according to the state value distribution hash table; wherein the sub-determination value P corresponding to the i-th monitoring state value i satisfies the following condition: wherein S i is the number of occurrences of the i-th monitoring state value in the numerical sequence, S sum is the total number of monitoring state values in the numerical sequence, μ w is the position distribution mean of the i-th monitoring state value in the numerical sequence, μ is the density of the position distribution of the i-th monitoring state value in the numerical sequence, P i 1 is the ratio of the position distribution mean and the density of the position distribution of the i-th monitoring state value, reflecting the credibility of the position distribution mean corresponding to each state. generating an abnormality determination value P corresponding to the current monitoring time according to the sub-determination values P1, P2 and P3 corresponding to the first type state values, the second type state values and the third type state values respectively; P satisfies the following condition: P = P3 + P2 - P1.

2. The ankylosing spondylitis integrated management system according to claim 1, characterized in that, The fusion processing further comprises: if P > Y1, generating an abnormality reminding information; Y1 is an abnormality threshold.

3. The spondyloarthritis comprehensive management system of claim 2, wherein, The fusion processing further comprises: if P < Y1, not generating an abnormality reminding information; Y1 is an abnormality threshold.

4. The spondyloarthritis comprehensive management system of claim 1, wherein, Further comprising: an information acquisition module; The information acquisition module is configured to acquire a feature observation sequence corresponding to each monitoring period, the feature observation sequence comprising attribute values of at least one observable dimension corresponding to a target event.

5. The spondyloarthritis comprehensive management system of claim 4, wherein, Further comprising a target hidden Markov model, which is in communication connection with the information acquisition module and the determination module respectively; The target hidden Markov model is configured to generate a monitoring state value corresponding to each historical monitoring period according to feature observation sequences of a plurality of continuous historical monitoring periods corresponding to a current monitoring time.

6. The spondyloarthritis comprehensive management system of claim 5, wherein, The monitoring state value is a stiffness index.

7. The ankylosing spondylitis comprehensive management system according to claim 6, characterized in that: if the stiffness index is in a range of [0, 1.3], the monitoring state value is a first type state value; if the stiffness index is in a range of (1.3, 2.1], the monitoring state value is a second type state value; if the stiffness index is in a range of (2.1, 10], the monitoring state value is a third type state value.

8. The ankylosing spondylitis integrated management system according to claim 5, characterized in that, The target hidden Markov model is acquired by the following method: collecting historical information of the target event in a plurality of continuous historical monitoring periods; the historical information comprises feature observation sequences and monitoring state values corresponding to the target event; training an initial hidden Markov model using the historical information of the target event, to generate the target hidden Markov model.

9. The spondylarthrities comprehensive management system according to claim 8, wherein, Further comprising a model updating module, which is configured to acquire feedback information of a user on accuracy of each abnormality prediction information generated according to the abnormality determination value; the feedback information comprises that the prediction information is accurate, the prediction information is too high, and the prediction information is too low; When the historical information of the target event in the plurality of continuous historical monitoring periods generated by the current user in the current update period is used to generate a user feature update training set; G1 is the number of feedback information that is accurate for prediction information obtained from the last update to the current period; G is the total number of feedback information obtained from the last update to the current period; Y2 is a model update threshold value; the monitoring state value Q in the user feature updating training set satisfies the following condition: Q = Q1 × K; Wherein, Q1 is the original abnormal prediction information corresponding to Q; K is the adjustment coefficient of the feedback information of the user to Q1, when the feedback information is accurate, K=1, when the feedback information is too high, K=0.8, when the feedback information is too low, K=1.2; The target hidden Markov model is updated and trained using the user feature update training set.

10. The ankylosing spondylitis integrated management system according to claim 4, characterized in that, The observable dimensions include pain and fatigue of the neck, back or hip.

Images