Device for recognizing Parkinson's gait disorder and computer readable storage medium
Through the local and global spatial feature fusion and multi-scale feature extraction of full pressure data, Parkinson's gait disorder is identified, which solves the problem that gait measurements in the prior art are difficult to capture subtle changes, and improves the accuracy of identification.
Patent Information
- Application Number
- CN202411400205.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-09
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-10-09
AI Technical Summary
When identifying Parkinson's gait disorders, it is difficult to capture subtle changes through statistical indicators of gait measurements, resulting in difficulty in early identification and monitoring, and short testing time, affecting accuracy.
Using full pressure data, local spatial features are extracted through the first model, global spatial features are extracted through the second model, and feature fusion is performed. Multi-scale feature extraction is used in combination with the original input data to obtain the spatiotemporal features of the gait to identify the Parkinson with gait disorder.
It improves the recognition accuracy of Parkinson's gait disorder, can extract gait information more comprehensively, and enhances the accuracy of early recognition and monitoring.
Smart Images

Figure CN118902435A_ABST
Abstract
Description
Technical Field
[0001] The present application generally relates to the technical field of Parkinson's gait detection, and more specifically, to a device and a computer-readable storage medium for identifying Parkinson's disease with gait disorder. Background Art
[0002] Parkinson's disease (PD) is a common neurodegenerative disease, among which postural gait disorders (PIGD), including panic gait and freezing gait, are one of the most important symptoms of PD and have the most serious impact on physical health. In the early stages of PD, as the disease progresses, gait variation will continue to increase, causing patients to frequently fall, get injured, and even fracture, seriously affecting their lives and health. Therefore, early identification of gait disorders in Parkinson's patients so that early treatment intervention can be carried out is of great significance to delaying the progression of Parkinson's disease and improving the quality of life of patients.
[0003] Plantar pressure gait analysis is a professional gait measurement method. By analyzing the data of the subjects walking on a walking device covered with pressure sensors (such as a gait blanket), the mechanical properties of the interaction between the feet and the ground when the individual is walking or standing are evaluated, thereby helping doctors capture and monitor individual gait changes. At present, a large number of studies are assisting the identification of PIGD by analyzing statistical indicators such as step length, pace, support time and swing time obtained by the subjects during gait measurement (see CN114999001A, CN116206765A). It can be understood that the aforementioned indicators are obtained based on the average value of the subjects' measurement data. Therefore, only the aforementioned statistical indicators are used for auxiliary identification, and subtle gait changes during the measurement of the subjects cannot be captured. In addition, in gait measurement, the gait blanket used is usually no more than 10 meters, and the time for the subjects to pass through the gait blanket is usually no more than 1 minute, which makes the test time short, and its statistical indicators such as step length and pace are difficult to change in a short time, so it is not conducive to the early identification and monitoring of PIGD.
[0004] In view of this, there is an urgent need to provide a solution for identifying Parkinson's disease with gait disorders. Parkinson's disease with gait disorders can be identified by using the full pressure data in gait detection. These full pressure data contain richer gait information and can capture more subtle changes in the subject's gait, thereby improving the accuracy of identifying Parkinson's disease with gait disorders. Summary of the invention
[0005] In order to at least solve one or more of the above-mentioned technical problems, the present application proposes a solution for identifying Parkinson's disease with gait disorder in multiple aspects.
[0006] In a first aspect, the present application provides a device for identifying Parkinson's disease with gait disorders, comprising: a processor; a memory, in which program instructions for identifying Parkinson's disease with gait disorders are stored, and when the program instructions are executed by the processor, the device implements the following operations: obtaining original input data, wherein the original input data at least includes full pressure data in gait measurement; based on the full pressure data, using a first model and a second model to extract local spatial features and global spatial features of gait respectively; performing feature fusion on the local spatial features and the global spatial features to obtain a fusion result; and based on the fusion result and the original input data, using a third model to perform multi-scale feature extraction to obtain spatiotemporal features of gait, so as to identify Parkinson's disease with gait disorders.
[0007] In some embodiments, the original input data further includes clinical information and gait statistical indicators, and the clinical information includes at least age, disease history or scale score information, and the gait statistical indicators include at least step length or step speed.
[0008] In other embodiments, when the program instructions are executed by the processor, the device further performs the following operations to extract local spatial features of gait: based on the full pressure data, using the first model to extract features between adjacent footprints of the subject in gait measurement to extract the local spatial features of gait.
[0009] In some further embodiments, when the program instructions are executed by the processor, the device further performs the following operations to extract the global spatial features of gait: based on the full pressure data, using the second model to extract the complete footprint features of the subject in the gait measurement to extract the global spatial features of gait.
[0010] In some further embodiments, the first model is a convolutional neural network model based on adaptive window attention.
[0011] In some further embodiments, the second model is a convolutional neural network model based on Fourier transform.
[0012] In some further embodiments, the third model includes a plurality of convolutional blocks of different scales.
[0013] In some other embodiments, when the program instructions are executed by the processor, the device further performs the following operations to obtain the spatiotemporal characteristics of gait: determining the time series of the subject's foot landing during gait measurement based on the fusion result and the original input data; inputting the time series into the third model, and extracting the time series features corresponding to different scales through the multiple convolution blocks of different scales; and fusing the time series features corresponding to different scales to obtain the spatiotemporal characteristics of gait.
[0014] In some further embodiments, when the program instructions are executed by the processor, the device further performs the following operations: inputting the spatiotemporal features into a fully connected layer, and identifying Parkinson's disease with gait disorder via an activation function layer to obtain an identification result of Parkinson's disease with gait disorder.
[0015] In a second aspect, the present application provides a computer-readable storage medium, characterized in that computer-readable instructions for identifying Parkinson's disease with gait disorder are stored thereon, and when the computer-readable instructions are executed by one or more processors, the operations performed by the device as described in one or more embodiments of the first aspect above are implemented.
[0016] Through the scheme for identifying Parkinson's disease with gait disorder provided above, the embodiment of the present application uses the full amount of pressure data in gait measurement, extracts the local spatial features and global spatial features of gait and fuses them, and then extracts multi-scale features based on the fusion results and the original input data to obtain the spatiotemporal features of gait, so as to identify Parkinson's disease with gait disorder. Based on this, it is possible to more comprehensively extract the rich gait information in the full amount of pressure data to improve the accuracy of Parkinson's disease with gait disorder. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become easy to understand. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 is an exemplary structural block diagram showing an apparatus for identifying Parkinson's disease with gait disorder according to an embodiment of the present application; Figure 2 is an exemplary flowchart showing operations performed by an apparatus for identifying Parkinson's disease with gait disorder according to an embodiment of the present application; Figure 3 is an exemplary schematic diagram showing a method for identifying Parkinson's disease with gait disorder according to an embodiment of the present application; Figure 4is an exemplary structural block diagram showing a device for identifying Parkinson's disease with gait disorder according to an embodiment of the present application. DETAILED DESCRIPTION
[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.
[0019] It should be understood that the terms "include" and "comprising" used in the specification and claims of the present application indicate the presence of described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.
[0020] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this application specification and claims, unless the context clearly indicates otherwise, the singular forms of "a", "an" and "the" are intended to include plural forms. It should also be further understood that the term "and / or" used in this application specification and claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.
[0021] As used in this specification and claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrases "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.
[0022] Figure 1 1 is an exemplary structural block diagram showing an apparatus 100 for identifying Parkinson's disease with gait disorder according to an embodiment of the present application. Figure 1As shown in , the device 100 may include a processor 101 and a memory 102. The processor 101 may include, for example, a general-purpose processor ("CPU") or a dedicated graphics processor ("GPU"), and the memory 102 stores program instructions that can be executed on the processor. In some embodiments, the memory 102 may include, but is not limited to, a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), and an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory).
[0023] Furthermore, the memory 102 may store program instructions for identifying Parkinson's disease with gait disorders. When the program instructions are executed by the processor, the device 100 performs the following operations: obtaining original input data, wherein the original input data at least includes full pressure data in gait measurement; based on the full pressure data, using a first model and a second model to extract local spatial features and global spatial features of the gait respectively; performing feature fusion on the local spatial features and the global spatial features to obtain a fusion result, and based on the fusion result and the original input data, using a third model to perform multi-scale feature extraction to obtain the spatiotemporal features of the gait, so as to identify Parkinson's disease with gait disorders.
[0024] As can be seen from the previous article, the above-mentioned full pressure data can be obtained by collecting data of Parkinson's patients and healthy subjects walking on a walking device (such as a gait blanket) covered with pressure sensors, which is the pressure information of all sensors on the walking device. In some implementation scenarios, the above-mentioned full pressure data can be represented by a four-dimensional feature vector P (t, x, y, p), where t represents time, x and y represent the coordinates of the pressure sensor on the gait blanket, and p represents the pressure value of the pressure sensor at time t.
[0025] In some embodiments, the above raw input data may also include clinical information and gait statistics, and the aforementioned clinical information may include but is not limited to the age, disease history or scale score information of Parkinson's patients and healthy subjects (such as H&Y, TUG, Berg balance and other scale score information). The aforementioned gait statistics may include but are not limited to step length, step speed, support time or swing time, etc. In some implementation scenarios, the aforementioned clinical information and gait statistics may be represented by a one-dimensional vector after, for example, ont-hot encoding.
[0026] Based on the original input data obtained as above, when the program instructions are executed by the processor, the device 100 further executes the extraction of local spatial features and global spatial features of gait based on the full pressure data using the first model and the second model, respectively. This is because at any time when the subject performs gait measurement, only the pressure sensor within the range of the subject's foot landing is activated. Therefore, the embodiment of the present application uses the first and second models to extract the spatial features of the subject's gait in gait detection in parallel, and the spatial features include local spatial features and global spatial features. By leveraging the complementary advantages of local and global spatial features, the system's ability to identify subtle gait changes in patients is improved.
[0027] For local spatial features, in some embodiments, the first model can be used to extract features between adjacent footprints of the subject in gait measurement based on the full pressure data to extract the local spatial features of the gait. In some implementation scenarios, the first model can be, for example, a convolutional neural network model based on adaptive window attention. Based on this, fine-grained spatial features can be extracted from the full pressure sensor data based on local convolution to capture more subtle changes in the subject's gait.
[0028] For global spatial features, in some embodiments, the second model can be used to extract the complete footprint features of the subject in gait measurement based on the full pressure data to extract the global spatial features of the gait. In some implementation scenarios, the second model can be, for example, a convolutional neural network model based on Fourier transform. That is, the global spatial features are extracted from the complete footprint information of the subject by Fourier transform method.
[0029] In other words, the convolutional neural network model based on the adaptive window attention mechanism can automatically identify the pressure characteristics between adjacent footprints of the subject, and then learn the differences between each step of the subject. The global feature extraction based on the global Fourier transform regards the entire measurement process of the subject as a whole for feature learning, promotes the fusion and complementarity of the subject's local gait information and global gait information, thereby improving the ability to identify subtle gait changes of patients and capturing more subtle changes in the subject's gait.
[0030] After extracting the local spatial features and global spatial features of the gait, the device 100 further performs feature fusion of the local spatial features and the global spatial features to obtain a fusion result. In some embodiments, the feature fusion is not limited to feature addition, splicing, weighted summation, etc. Through the aforementioned feature fusion, the complementary advantages between the complete footprint features of the patient and the subtle local features between adjacent footprints can be comprehensively utilized to improve the robustness of the system.
[0031] It can be understood that human walking movement depends largely on the changes in the patient's footprint over time. Therefore, the above-mentioned device 100 of the embodiment of the present application further performs multi-scale feature extraction based on the fusion results and the original input data using a third model to obtain the spatiotemporal characteristics of the gait. That is, by inputting the extracted spatial feature vector together with the original input into the third model, the spatiotemporal characteristics of the gait are obtained through multi-scale feature extraction. In some embodiments, the aforementioned third model includes a plurality of convolution blocks of different scales. As an example, the third model may include three convolution blocks with convolution kernel sizes of 1×1, 3×3 and 7×7. Based on this, the patient's gait information can be extracted from the perspective of timing changes to further improve the accuracy and robustness of the system.
[0032] In some embodiments, the time series of the subject's foot landing in the gait measurement is first determined based on the fusion result and the original input data, and then the time series is input into the third model, and the time series features corresponding to the different scales are respectively extracted through a plurality of convolution blocks of different scales, and then the time series features corresponding to the different scales are fused to obtain the spatiotemporal features of the gait. In some implementation scenarios, the aforementioned time series features include the pressure value of the sensor. It can be understood that during the gait measurement process, the pressure value of the sensor is not 0 only when the subject's foot lands. Thus, the time series of the subject's foot landing can be extracted from the entire gait measurement data, and the pressure value of the pressure sensor at each landing is used as the input of the third model, and the features at different scales are extracted and fused. Repeat the above process until the time series information of all landing moments is extracted, and the spatiotemporal features of the subject's gait are obtained. In some implementation scenarios, the feature fusion at different scales can be implemented by, for example, summing or weighted summing.
[0033] Based on the extracted spatiotemporal features of gait, the device 100 further inputs the spatiotemporal features into a fully connected layer, and identifies Parkinson's disease with gait disorder via an activation function layer to obtain an identification result of Parkinson's disease with gait disorder, wherein the identification result is the probability of whether the subject has Parkinson's disease with gait disorder.
[0034] Combined with the above description, it can be seen that the embodiment of the present application uses the full amount of pressure data in the gait measurement, and extracts the local spatial features and global spatial features of the gait in parallel through the first model (for example, a convolutional neural network model based on adaptive window attention) and the second model (for example, a convolutional neural network model based on Fourier transform) and fuses them, and then performs multi-scale feature extraction based on the fusion result and the original input data third model to obtain the spatiotemporal features of the gait to identify Parkinson's disease with gait disorders. Based on this, it is possible to more comprehensively extract rich gait information in the full amount of pressure data to improve the accuracy of Parkinson's disease with gait disorders.
[0035] Figure 2 FIG. 2 is an exemplary flowchart showing the operation 200 performed by the apparatus for identifying Parkinson's disease with gait disorder according to an embodiment of the present application. Figure 2 As shown in, at step S201, the original input data is obtained. In some embodiments, the original input data may include, for example, full pressure data, clinical information, and gait statistical indicators. According to the foregoing, the aforementioned full pressure data can be obtained by collecting data of Parkinson's patients and healthy subjects walking on a walking device (such as a gait blanket) covered with pressure sensors. The aforementioned clinical information may include the age, disease history, H&Y, TUG, Berg balance, etc. of Parkinson's patients and healthy subjects. The aforementioned gait statistical indicators may include step length, walking speed, support time or swing time, etc.
[0036] Next, at step S202, based on the full amount of pressure data, the local spatial features of the gait are extracted using the first model. In some embodiments, the first model may be, for example, a convolutional neural network model based on adaptive window attention. Specifically, the features between adjacent footprints of the subject in the gait measurement are extracted by using the first model to extract the local spatial features of the gait. In addition, at step S203, based on the full amount of pressure data, the global spatial features of the gait are extracted using the second model. In some embodiments, the second model may be, for example, a convolutional neural network model based on Fourier transform. Specifically, the complete footprint features of the subject in the gait measurement are extracted by using the second model to extract the global spatial features of the gait.
[0037] After extracting the local spatial features and global spatial features of the gait, at step S204, the local spatial features and the global spatial features are feature fused to obtain a fusion result. In some embodiments, the local spatial features and the global spatial features can be feature fused to obtain a fusion result by, for example, feature addition, splicing, weighted summation, etc. Further, at step S205, based on the fusion result and the original input number, a third model is used to extract time series features corresponding to different scales using multi-scale features. In some embodiments, the aforementioned third model may include convolution blocks of multiple different scales (e.g., 1×1, 3×3, and 7×7). In this scenario, the time series of the subject's foot landing during gait measurement is first determined based on the fusion result and the original input data, and then the time series is input into the third model, and the time series features corresponding to different scales are respectively extracted through convolution blocks of multiple different scales. Further, at step S206, the time series features corresponding to different scales are fused to obtain the spatiotemporal features of the gait. Similar to the feature fusion of the local features and the global features described above, the temporal features corresponding to different scales may also be fused by, for example, summing or weighted summing.
[0038] The figure further shows that at step S207, by inputting the spatiotemporal features into the fully connected layer and identifying Parkinson's disease with gait disorder via the activation function layer, the identification result of Parkinson's disease with gait disorder can be obtained. Based on this, the rich gait information in the full amount of pressure data can be more comprehensively extracted to improve the accuracy of Parkinson's disease with gait disorder.
[0039] Figure 3 FIG. 2 is an exemplary schematic diagram showing a method for identifying Parkinson's disease with gait disorder according to an embodiment of the present application. Figure 3 As shown in , the acquired full amount of pressure data 301 is input into the first model 302 and the second model 303 in parallel. As mentioned above, the first model can be, for example, a convolutional neural network model based on adaptive window attention, and the features between adjacent footprints of the subject in gait measurement are extracted by using the first model to extract the local spatial features 304 of the gait. The aforementioned second model can be, for example, a convolutional neural network model based on Fourier transform, and the complete footprint features of the subject in gait measurement are extracted by using the second model to extract the global spatial features 305 of the gait. Then, the local spatial features and the global spatial features are feature fused to obtain the fusion result 306.
[0040] Furthermore, the aforementioned fusion result 306 and the original input data (taking the full pressure data 301 as an example) are input into the third model 307. The figure exemplarily shows that the third model 307 includes three convolution blocks 307-1, 307-2 and 307-3 with convolution kernel sizes of 1×1, 3×3 and 7×7 respectively. Specifically, firstly, based on the fusion result and the original input data, the time series 308 of the subject when the foot lands in the gait measurement is determined, and the time series features corresponding to different scales are extracted from the time series 308 via the convolution blocks 307-1, 307-2 and 307-3 respectively. By, for example, adding (for example, the “ ” as shown in the figure, the time series features corresponding to different scales are fused to obtain the spatiotemporal features of gait 309. Finally, the spatiotemporal features 309 are input into the fully connected layer 310, and the Parkinson's disease with gait disorder is identified through the activation function (such as softmaxt) 311 to obtain the recognition result 312 of Parkinson's disease with gait disorder. The recognition result is the probability of whether the subject has Parkinson's disease with gait disorder.
[0041] Figure 4 4 is an exemplary structural block diagram of a device 400 for identifying Parkinson's disease with gait disorder according to an embodiment of the present application. It is understood that the device 400 may include the apparatus of the embodiment of the present application, and the device implementing the solution of the present application may be a single device (such as a computing device) or a multifunctional device including various peripheral devices.
[0042] like Figure 4 As shown in , the device of the present application may also include a central processing unit or central processing unit ("CPU") 411, which may be a general-purpose CPU, a dedicated CPU, or other information processing and program execution unit. Further, the device 400 may also include a large-capacity memory 412 and a read-only memory ("ROM") 413, wherein the large-capacity memory 412 may be configured to store various types of data, including various raw input data, full pressure data, clinical information and gait statistics, algorithm data, intermediate results, and various programs required to run the device 400. ROM 413 may be configured to store the power-on self-test of the device 400, the initialization of each functional module in the system, the basic input / output driver of the system, and the data and instructions required to boot the operating system.
[0043] Optionally, the device 400 may also include other hardware platforms or components, such as the tensor processing unit ("TPU") 414, the graphics processing unit ("GPU") 415, the field programmable gate array ("FPGA") 416, and the machine learning unit ("MLU") 417 shown. It is understood that although a variety of hardware platforms or components are shown in the device 400, they are only exemplary and not restrictive, and those skilled in the art may add or remove corresponding hardware according to actual needs. For example, the device 400 may only include a CPU, related storage devices, and interface devices to implement the operations performed by the present application for identifying Parkinson's disease with gait disorders.
[0044] In some embodiments, in order to facilitate the transmission and interaction of data with an external network, the device 400 of the present application further includes a communication interface 418, so that it can be connected to a local area network / wireless local area network ("LAN / WLAN") 405 through the communication interface 418, and then connected to a local server 406 or to the Internet ("Internet") 407 through the LAN / WLAN. Alternatively or additionally, the device 400 of the present application can also be directly connected to the Internet or a cellular network through the communication interface 418 based on wireless communication technology, such as wireless communication technology based on the third generation ("3G"), the fourth generation ("4G") or the fifth generation ("5G") generation. In some application scenarios, the device 400 of the present application can also access a server 408 and a database 409 of an external network as needed to obtain various known algorithms, data and modules, and can remotely store various data, such as various types of data or instructions for presenting, for example, full pressure data, clinical information and gait statistics.
[0045] The peripheral devices of the device 400 may include a display device 402, an input device 403, and a data transmission interface 404. In one embodiment, the display device 402 may include, for example, one or more speakers and / or one or more visual displays, which are configured to provide voice prompts and / or image video displays for the present application for identifying Parkinson's disease with gait disorders. The input device 403 may include other input buttons or controls such as a keyboard, a mouse, a microphone, a gesture capture camera, etc., which are configured to receive input of audio data and / or user instructions. The data transmission interface 404 may include, for example, a serial interface, a parallel interface or a universal serial bus interface ("USB"), a small computer system interface ("SCSI"), a serial ATA, a FireWire ("FireWire"), a PCI Express, and a high-definition multimedia interface ("HDMI"), etc., which are configured for data transmission and interaction with other devices or systems. According to the solution of the present application, the data transmission interface 404 can receive the full amount of pressure data collected from the walking collection device and the clinical information and gait statistics in the medical database, and transmit the full amount of pressure data, clinical information and gait statistics or various other types of data or results to the device 400.
[0046] The CPU 411, mass storage 412, ROM 413, TPU 414, GPU 415, FPGA 416, MLU 417 and communication interface 418 of the device 400 of the present application can be interconnected through a bus 419, and data interaction with peripheral devices can be achieved through the bus. In one embodiment, through the bus 419, the CPU 411 can control other hardware components in the device 400 and its peripheral devices.
[0047] Combination of the above Figure 4 The present invention describes a device for identifying Parkinson's disease with gait disorders that can be used to implement the present invention. It should be understood that the device structure or architecture here is only exemplary, and the implementation method and implementation entity of the present invention are not limited thereto, but can be changed without departing from the spirit of the present invention.
[0048] According to the above description in combination with the accompanying drawings, those skilled in the art can also understand that the embodiments of the present application can also be implemented by software programs. Therefore, the present application also provides a computer-readable storage medium, on which computer-readable instructions for identifying Parkinson's disease with gait disorder are stored. When the computer-readable instructions are executed by one or more processors, they can be used to implement the present application in combination with the accompanying drawings. Figure 2 Operations performed by the described apparatus for identifying Parkinson's disease associated with gait disturbance.
[0049] It should be noted that although the operations of the method of the present application are described in a specific order in the accompanying drawings, this does not require or imply that the operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired results. On the contrary, the steps depicted in the flow chart can be performed in a different order. Additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step, and / or one step can be decomposed into multiple steps.
[0050] It should be understood that when the terms "first", "second", "third" and "fourth" are used in the claims, the specification and the drawings of the present application, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections.
[0051] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this application specification and claims, unless the context clearly indicates otherwise, the singular forms of "a", "an" and "the" are intended to include plural forms. It should also be further understood that the term "and / or" used in this application specification and claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.
[0052] Although the implementation methods of the present application are as above, the contents described are only examples adopted to facilitate the understanding of the present application, and are not intended to limit the scope and application scenarios of the present application. Any technician in the technical field described in the present application can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present application, but the scope of patent protection of the present application shall still be subject to the scope defined in the attached claims.
[0053] In addition, the collection and acquisition of various data in this application complies with relevant laws and regulations and is authorized by the data provider. Any organization or individual that needs to obtain external data must obtain authorization in accordance with the law and ensure data security. It is not allowed to illegally collect, use, process, or transmit unauthorized or unprotected data, or to illegally buy, sell, provide, or disclose unauthorized or unprotected data.
Claims
1. A device for identifying Parkinson's disease with gait disorder, It is characterized in that include: processor; A memory having program instructions for identifying Parkinson's disease with gait disorder stored therein, wherein when the program instructions are executed by the processor, the device implements the following operations: Acquiring original input data, wherein the original input data at least includes full pressure data in gait measurement; Based on the full pressure data, using the first model and the second model to extract local spatial features and global spatial features of gait respectively; Performing feature fusion on the local spatial feature and the global spatial feature to obtain a fusion result; as well as Based on the fusion result and the original input data, a third model is used to perform multi-scale feature extraction to obtain the spatiotemporal characteristics of gait, so as to identify Parkinson's disease with gait disorder.
2. The device according to claim 1, It is characterized in that The original input data further includes clinical information and gait statistical indicators, and the clinical information at least includes age, disease history or scale score information, and the gait statistical indicators at least include step length or step speed.
3. The device according to claim 1, It is characterized in that When the program instructions are executed by the processor, the device further performs the following operations to extract local spatial features of gait: Based on the total pressure data, the first model is used to extract features between adjacent footprints of the subject in gait measurement, so as to extract the local spatial features of gait.
4. The device according to claim 1, It is characterized in that When the program instructions are executed by the processor, the device further performs the following operations to extract the global spatial features of the gait: Based on the full pressure data, the second model is used to extract complete footprint features of the subject in gait measurement to extract the global spatial features of gait.
5. The device according to claim 1 or 3, It is characterized in that The first model is a convolutional neural network model based on adaptive window attention.
6. The device according to claim 1 or 4, It is characterized in that The second model is a convolutional neural network model based on Fourier transform.
7. The device according to claim 1, It is characterized in that The third model includes multiple convolution blocks of different scales.
8. The device according to claim 7, It is characterized in that When the program instructions are executed by the processor, the device further performs the following operations to obtain the spatiotemporal characteristics of gait: Determine the time series of the subject's foot landing during gait measurement based on the fusion result and the original input data; Inputting the time series into the third model, and extracting time series features corresponding to different scales through the multiple convolution blocks of different scales respectively; as well as The temporal features corresponding to the different scales are fused to obtain the spatiotemporal features of the gait.
9. The device according to claim 1, It is characterized in that When the program instructions are executed by the processor, the device further performs the following operations: The spatiotemporal features are input into the fully connected layer, and Parkinson's disease with gait disorder is identified through the activation function layer to obtain the identification result of Parkinson's disease with gait disorder.
10. A computer-readable storage medium, It is characterized in that Computer-readable instructions for identifying Parkinson's disease with gait disorder are stored thereon, and when the computer-readable instructions are executed by one or more processors, the operations performed by the device as described in any one of claims 1 to 9 are implemented.
Citation Information
Patent Citations
Parkinson patient walking ability evaluation method based on gait time-space parameters and three-dimensional force characteristics
CN104598722A
assessment method and system for Parkinson's disease gait dyskinesia severity, and equipment
CN111382679A
Abnormal gait behavior detection system and method, electronic equipment and readable storage medium
CN115517666A
Device for predicting cognitive impairment and computer readable storage medium
CN115590481A
Device for predicting cognitive impairment of Parkinson's disease and storage medium
CN116205909A
Cited By
Parkinson's disease gait monitoring equipment and working method thereof
CN122163203A
