CN115886787B - Ground reaction force transformation method for disease screening, bone disease screening system and equipment - Google Patents
Ground reaction force transformation method for disease screening, bone disease screening system and equipment Download PDFInfo
- Publication number
- CN115886787B CN115886787B CN202310221888.0A CN202310221888A CN115886787B CN 115886787 B CN115886787 B CN 115886787B CN 202310221888 A CN202310221888 A CN 202310221888A CN 115886787 B CN115886787 B CN 115886787B
- Authority
- CN
- China
- Prior art keywords
- disease screening
- pixel points
- reaction force
- grf
- ground reaction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
A ground reaction force transformation method for disease screening, a bone disease screening system and equipment belong to the technical field of human motion analysis. The three-dimensional ground reaction force detection method aims at solving the problems that the existing three-dimensional ground reaction force for disease screening has more lost information, the correlation among different pressure platform characteristic information is weak, and the like. The invention relates to a ground reaction force conversion method for disease screening, which comprises the steps of firstly converting three numerical values of GRF (gradient) of each frame in a left pressure platform and a right pressure platform of a group of pressure platforms in the vertical direction, the front-rear direction and the inner-outer direction into R, G, B numerical values in RGB images to obtain pixel points corresponding to GRF data of the two pressure platforms, and simultaneously carrying out correlation processing on GRFs of frames corresponding to the two pressure platforms to obtain pixel points reflecting the correlation; and synthesizing a color picture according to R, G, B pixel points corresponding to each frame GRF in the left pressure platform and the right pressure platform. The invention is suitable for screening bone diseases.
Description
Technical Field
The invention belongs to the technical field of human body movement analysis, and particularly relates to a ground reaction force transformation method, a bone disease screening system and equipment.
Background
In common orthopedic diseases, patients often show gait abnormalities to varying degrees after bone and joint and muscle ligaments are damaged. The three-dimensional ground reaction force (Ground Reaction Force, GRF) can measure the stress condition of the body in three dimensions in the walking process of the patient, and can reflect the abnormal body functions of the patient to a certain extent. Compared with the traditional gait analysis based on motion capture, the GRF measurement and analysis are simple and quick, do not need extra preparation time, and are therefore suitable for clinical rapid examination.
GRF, however, measures a continuous pressure value curve of the foot on the pressure platform during walking. How to extract effective features for disease screening and diagnosis is a key and difficult point of disease screening based on GRF. Common features include loading rate of the GRF in the vertical direction, first peak and second peak, etc.; the characteristics of the GRF in the front-rear direction such as the maximum braking force and the maximum propelling force, and the characteristic parameters of the correlation between different pressure platforms which can be obtained by further processing the characteristics directly extracted by the GRF comprise symmetry angles, symmetry indexes and the like. But the above features are local features extracted based on the complete GRF curve, losing most of the global features. Thus, some important information may be ignored. When the classification screening is carried out by adopting a common machine learning model and the like according to the extracted characteristic parameters, the overall screening accuracy is not high. Therefore, how to perform disease screening according to complete ground reaction force information and correlation information between different pressure platforms is one of the key factors for improving the accuracy of GRF-based screening.
Disclosure of Invention
The invention aims to solve the problems that the existing three-dimensional ground reaction force for disease screening has more lost information, and the correlation among the characteristic information of different pressure platforms is weak.
The ground reaction force conversion method for disease screening includes the steps of firstly converting three numerical values of GRF of each frame in a left pressure platform and a right pressure platform of a group of pressure platforms in the vertical direction, the front-rear direction and the inner-outer direction into R, G, B numerical values in RGB images to obtain pixel points corresponding to GRF data of the two pressure platforms, and performing correlation processing on GRFs of corresponding frames of the two pressure platforms to obtain pixel points reflecting the correlation; and synthesizing a color picture according to R, G, B pixel points corresponding to each frame GRF in the left pressure platform and the right pressure platform.
Further, the manner of performing the correlation processing includes the following processing manners:
and solving maximum, minimum, average and difference values of GRF data of the frames corresponding to the two pressure platforms.
Further, in the process of carrying out correlation processing on GRFs of frames corresponding to the two pressure platforms, two correlation processing modes are adopted to carry out correlation processing on GRF data of the two pressure platforms, so that two groups of pixel points reflecting the correlation are obtained.
Further, the process of synthesizing the color picture according to R, G, B pixel points corresponding to each frame GRF in the left pressure platform and the right pressure platform comprises the following steps:
s21, respectively marking the pixel points of the two pressure platforms as GA and GB, wherein the number of the pixel points of the GA and GB is N; two groups of pixel points reflecting the correlation obtained by the two correlation processing modes are respectively marked as GABA and GABB, and the number of the GABA and GABB pixel points is N;
reconstructing N one-dimensional pixel points of GA, GB, GABA, GABB into (N/M) multiplied by M two-dimensional pixel points respectively;
and S22, splicing the two-dimensional pixel points corresponding to GA, GB, GABA, GABB to form square two-dimensional pixel points of M multiplied by M, so as to form a color picture.
Further, the manner of stitching the two-dimensional pixel points corresponding to GA, GB, GABA, GABB includes: directly splicing GA, GB, GABA, GABB, and/or cross-splicing GA, GB, GABA, GABB.
Further, before converting the values of the GRF in three directions into RGB values, it is necessary to perform a normalization operation on the GRF values, and then convert the data after the normalization operation into values of R, G, B three channels of the RGB image;
the normalization operation comprises the following steps:
firstly, recording the resultant force peak-valley values of GRF in three directions as PV_R, the vertical peak-valley value PV_V, the front-rear peak-valley value PV_AP and the internal-external peak-valley value PV_ML; according to the min-max standardization method, the vertical direction, the front-back direction and the inner-outer direction of the GRF curve are respectively normalized to [0, PV_V/PV_R ], [0, PV_AP/PV_R ], [0, PV_ML/PV_R ].
A ground reaction force conversion apparatus for disease screening, the apparatus comprising a processor and a memory having stored therein at least one instruction loaded and executed by the processor to implement a ground reaction force conversion method for disease screening.
The bone disease screening system comprises a ground reaction force image acquisition module and a bone disease screening module;
the ground reaction force image acquisition module: the method is used for obtaining a color picture obtained based on a ground reaction force transformation method for disease screening;
bone disease screening module: taking the color picture acquired by the ground reaction force image acquisition module as input, and calling a bone disease screening model to classify bone diseases;
the bone disease screening model is a machine learning model; the training process of the bone disease screening model comprises the following steps of:
obtaining a color picture by using a ground reaction force transformation method for disease screening, and constructing a data set; and training the bone disease screening model by using the constructed data set to obtain a trained bone disease screening model.
Further, the bone disease screening model adopts a Resnet classification model.
A bone screening device comprising a processor and a memory, wherein the memory stores at least one instruction that is loaded and executed by the processor to implement a bone screening system.
The beneficial effects are that:
the invention provides a conversion method for converting the ground reaction curve characteristics into the picture information, which can completely reserve the global information characteristics of GRFs, can increase the associated information characteristics among GRFs of different pressure platforms, further utilizes a classification model in more mature machine learning to screen diseases, and can effectively improve the accuracy rate of disease screening. Therefore, the invention not only can effectively solve the problems of more lost information, weak relevance among the characteristic information of different pressure platforms and the like when the Ground Reaction Force (GRF) is utilized to carry out characteristic extraction, but also can solve the problem of low accuracy of a machine learning model for screening diseases by utilizing the traditional GRF characteristics.
Meanwhile, the method generates smaller pictures, the screening speed can reach the second level, and the method has larger application potential.
Drawings
Fig. 1 is a graph of the vertical magnitude of the left pressure plateau.
Fig. 2 is a graph of the vertical magnitude of the right pressure plateau.
Fig. 3 is a graph showing the numerical value of the left side pressure plateau in the front-rear direction.
Fig. 4 is a graph showing the numerical value of the right side pressure plateau in the front-rear direction.
Fig. 5 is a graph of the magnitude of the left pressure plateau in the inward and outward directions.
Fig. 6 is a graph of the magnitude of the right side pressure in the inward and outward directions of the plateau.
Fig. 7 is an image of the ground reaction force configuration of the pressure table data.
FIG. 8 is a schematic view of the walkway and pressure platform placement.
Fig. 9 is a corresponding classification result of screening for scoliosis in an implementation.
Detailed Description
The first embodiment is as follows:
the present embodiment is a ground reaction force conversion method for disease screening. The invention adopts at least one group of pressure platforms for data acquisition, and one group of pressure platforms comprises a left pressure platform and a right pressure platform, and the left pressure platform and the right pressure platform acquire ground reaction force at the same time. In this embodiment, data collection using a set of pressure platforms is described as an example.
The ground reaction force transformation method for disease screening according to the embodiment comprises the following steps:
s1, converting three numerical values of the vertical direction, the front-back direction and the inner-outer direction of each frame GRF in the left pressure platform and the right pressure platform of a group of pressure platforms into R, G, B numerical values in an RGB image;
s2, synthesizing a color picture according to R, G, B values corresponding to each frame GRF in the left pressure platform and the right pressure platform.
In step S1, three values of the vertical direction, the front-rear direction and the inner-outer direction of the GRF in the present embodiment are shown in fig. 1 to 6, wherein fig. 1, 3 and 5 are three values of the vertical direction, the front-rear direction and the inner-outer direction of the left side pressure platform, and fig. 2, 4 and 6 are three values of the vertical direction, the front-rear direction and the inner-outer direction of the right side pressure platform;
the gait cycle shown on the abscissa in fig. 1 to 6 is 100, so the GRF data of each pressure platform can constitute 100 pixels, and the 100 pixels of the two pressure platforms are denoted as GA and GB, respectively;
correlation processing is performed on two pressure platforms to obtain correlation pixel points, and the method for performing correlation processing includes but is not limited to: and solving maximum, minimum, average, difference and the like of GRF data of frames corresponding to the two pressure platforms. The correlation processing is performed in at least two ways, and in practice the present invention may not be limited to two ways, and may be selected in plural as long as it is ensured that the subsequent synthesized color image is a square image, the subsequent color image may be 20×20 when two correlation processing ways are selected, the subsequent color image may be 30×30 when plural correlation processing ways are selected here, and so on.
In this embodiment, two correlation processes are adopted to obtain two groups of correlation pixel points, that is, two 100 pixel points reflecting the correlation of GA and GB are constructed according to GA and GB, and the two correlation pixel points are respectively marked as GABA and GABB (obtained by two modes of maximum, minimum, mean and difference);
before converting the GRF three-direction values into RGB values, the GRF values need to be normalized, and the normalization operation comprises the following steps:
before normalization, the peak-valley values of the resultant force of GRF in three directions are recorded as PV_R, the peak-valley value PV_V in the vertical direction, the peak-valley value PV_AP in the front-back direction and the peak-valley value PV_ML in the inside-outside direction. According to the min-max standardization method, the vertical direction, the front-back direction and the inside-outside direction of the GRF curve are respectively normalized to the values of [0, PV_V/PV_R ], [0, PV_AP/PV_R ], [0, PV_ML/PV_R ], and then are respectively used as the values of R, G, B channels of the RGB image.
The present embodiment uses the min-max normalization method, but the normalization method in the present invention is not limited to the min-max method.
In step S2, GA, GB, GABA, GABB has 400 color pixels, so that a 20×20 color image (as shown in fig. 7) can be formed, and the specific process includes the following steps:
s21, respectively reconstructing 100 one-dimensional pixel points of GA, GB, GABA, GABB into 5 multiplied by 20 two-dimensional pixel points;
and S22, splicing the two-dimensional pixel points corresponding to GA, GB, GABA, GABB to form square two-dimensional pixel points of 20 multiplied by 20, so as to form a color picture. The splicing manner in this embodiment may be any one of the following manners, or two manners may be combined:
(A) Directly splicing GA, GB, GABA, GABB to form a square two-dimensional pixel point with the size of 20 multiplied by 20, and forming a color picture. And respectively arranging and combining the splicing sequences of GA, GB, GABA and GABB to obtain 24 corresponding color pictures (such as the first 24 pictures in FIG. 7).
(B) Performing cross stitching on GA, GB, GABA, GABB, namely, 1×20 pixels in the first row of GA, GB, GABA, GABB form 1-4 rows of pixel points in a 20×20 picture; GA. The second row of 1×20 pixels of GB, GABA, GABB constitutes the first 5-8 rows of pixels in the 20×20 picture, and so on to constitute the final 20×20 color picture. GA, GB, GABA, GABB are respectively arranged and combined to obtain another group of 24 corresponding color pictures (such as 24 pictures after fig. 7).
The invention includes, but is not limited to, the above two splicing modes.
The ground reaction force conversion method for disease screening can effectively and completely reserve the global information characteristics of GRFs, can increase the associated information characteristics among GRFs of different pressure platforms, and provides an effective and automatic screening basis for the screening of the orthopedic diseases so as to realize rapid and accurate screening of the orthopedic diseases.
The second embodiment is as follows:
this embodiment is a ground reaction force conversion device for disease screening, the device including a processor and a memory, it should be understood that the device includes any device provided with a processor and a memory described in the present invention, and the device may also include other units, modules for performing display, interaction, processing, control and other functions through signals or instructions;
the memory stores at least one instruction that is loaded and executed by the processor to implement the ground reaction force conversion method for disease screening.
The memory may include a non-transitory machine-readable medium having instructions stored thereon, including, but not limited to, magnetic storage media, optical storage media; the magneto-optical storage medium includes: read only memory ROM, random access memory RAM, erasable programmable memory (e.g., EPROM and EEPROM), and flash memory layers; or other type of medium suitable for storing electronic instructions.
And a third specific embodiment:
this embodiment is a bone disease screening system, including ground counter-force image acquisition module and bone disease screening module:
the ground reaction force image acquisition module: the method is used for obtaining a color picture obtained by the ground reaction force transformation method for disease screening according to the first embodiment;
bone disease screening module: taking the color picture acquired by the ground reaction force image acquisition module as input, and calling a bone disease screening model to classify bone diseases;
the bone disease screening model is a machine learning model, and the machine learning model comprises but is not limited to a neural network model;
the bone disease screening model in this embodiment adopts a Resnet classification model. After a Resnet classification model is built, a color picture is obtained by using the ground reaction force transformation method for disease screening according to the first embodiment, and a data set is built; training a machine learning model by utilizing the data set; in the process, dividing the pictures into a healthy group and a disease group; of course, the present embodiment is classified into a healthy group and a disease group, but the grouping of the present invention is not limited to the classification into a healthy group and a disease group 2, and may be a variety of groupings including bone disease screening results.
And then screening bone diseases by using the trained image classification model, obtaining transformed color pictures by using the ground reaction force transformation method for disease screening according to the first embodiment, wherein each patient GRF can generate 48 color pictures, and disease screening can be performed according to any picture. And further screening and classifying 48 color pictures of each subject by adopting an image classification model, and taking the result with the largest occurrence number of the screening result in the 48 pictures as the final screening result of the current subject by adopting a few rules obeying majority.
The specific embodiment IV is as follows:
the embodiment is a bone disease screening device, which includes a processor and a memory, and should be understood to include any device provided with a processor and a memory described in the present invention, where the device may further include other units and modules that perform display, interaction, processing, control, and other functions through signals or instructions;
the memory stores at least one instruction that is loaded and executed by the processor to implement the bone screening system.
The memory may include a non-transitory machine-readable medium having instructions stored thereon, including, but not limited to, magnetic storage media, optical storage media; the magneto-optical storage medium includes: read only memory ROM, random access memory RAM, erasable programmable memory (e.g., EPROM and EEPROM), and flash memory layers; or other type of medium suitable for storing electronic instructions.
Examples
This example illustrates the detection of adolescent idiopathic scoliosis (adolescent idiopathic scoliosis, AIS) patients who walk back and forth naturally on the walkway for 5 minutes without informing the walkway of a pressure platform, and collect and record the ground reaction data for that 5 minutes, taking two AMTI three-dimensional force platforms (as shown in fig. 8, the two square marked positions in fig. 8 are two AMTI three-dimensional force platforms, the left and right two AMTI three-dimensional force platforms are not perfectly aligned, the left and right are relative to the left and right sides of a person walking). The mean curve of GRF in three dimensions during AIS patient walking was then calculated. AIS screening is carried out by adopting the GRF processing mode and the Resnet18 model, a screening classification result is schematically shown in fig. 9, AIS Patent in fig. 9 represents adolescent idiopathic scoliosis, and Health Subject represents the Health of a Subject; AIS Patient (AIS Patient) indicates actual and predicted adolescent idiopathic scoliosis, AIS Patient (Health Subject) indicates actual and predicted adolescent idiopathic scoliosis; health subjects (Health subjects) indicate actual and predicted to be healthy in subjects, and Health subjects (AIS events) indicate actual and predicted to be healthy in subjects for adolescent idiopathic scoliosis. By adopting the method, the screening effect with the accuracy of 82% can be realized after reasonable sample training.
The above examples of the present invention are only for describing the calculation model and calculation flow of the present invention in detail, and are not limiting of the embodiments of the present invention. Other variations and modifications of the above description will be apparent to those of ordinary skill in the art, and it is not intended to be exhaustive of all embodiments, all of which are within the scope of the invention.
Claims (6)
1. The ground reaction force conversion method for disease screening is characterized in that three values of the vertical direction, the front-back direction and the inner-outer direction of GRF of each frame in a group of pressure platforms are converted into R, G, B values in RGB images to obtain pixel points corresponding to GRF data of the two pressure platforms, and correlation processing is carried out on the GRFs of the frames corresponding to the two pressure platforms to obtain pixel points reflecting the correlation; then synthesizing a color picture according to R, G, B pixel points corresponding to each frame GRF in the left pressure platform and the right pressure platform;
the method for carrying out correlation processing comprises the following processing modes:
solving maximum, minimum, mean and difference of GRF data of corresponding frames of the two pressure platforms;
in the process of carrying out correlation processing on GRFs of frames corresponding to the two pressure platforms, adopting two correlation processing modes to carry out correlation processing on GRF data of the two pressure platforms to obtain two groups of pixel points reflecting the correlation;
the process of synthesizing the color picture according to R, G, B pixel points corresponding to each frame GRF in the left pressure platform and the right pressure platform comprises the following steps:
s21, respectively marking the pixel points of the two pressure platforms as GA and GB, wherein the number of the pixel points of the GA and GB is N; two groups of pixel points reflecting the correlation obtained by the two correlation processing modes are respectively marked as GABA and GABB, and the number of the GABA and GABB pixel points is N;
reconstructing N one-dimensional pixel points of GA, GB, GABA, GABB into (N/M) multiplied by M two-dimensional pixel points respectively;
and S22, splicing the two-dimensional pixel points corresponding to GA, GB, GABA, GABB to form square two-dimensional pixel points of M multiplied by M, so as to form a color picture.
2. The ground reaction force transformation method for disease screening according to claim 1, wherein the manner of stitching the two-dimensional pixel points corresponding to GA, GB, GABA, GABB comprises: directly splicing GA, GB, GABA, GABB, and/or cross-splicing GA, GB, GABA, GABB.
3. A ground reaction force conversion method for disease screening according to claim 1 or 2, wherein before converting the values of the GRF three directions into RGB values, it is necessary to perform a normalization operation on the GRF values, and then convert the data after the normalization operation into values of R, G, B three channels of RGB images;
the normalization operation comprises the following steps:
firstly, recording the resultant force peak-valley values of GRF in three directions as PV_R, the vertical peak-valley value PV_V, the front-rear peak-valley value PV_AP and the internal-external peak-valley value PV_ML; according to the min-max standardization method, the vertical direction, the front-back direction and the inner-outer direction of the GRF curve are respectively normalized to [0, PV_V/PV_R ], [0, PV_AP/PV_R ], [0, PV_ML/PV_R ].
4. A ground reaction force conversion device for disease screening, characterized in that the device comprises a processor and a memory, the memory having stored therein at least one instruction which is loaded and executed by the processor to implement a ground reaction force conversion method for disease screening as claimed in any one of claims 1 to 3.
5. The bone disease screening system is characterized by comprising a ground reaction force image acquisition module and a bone disease screening module;
the ground reaction force image acquisition module: a method for obtaining a color picture obtained based on the ground reaction force conversion method for disease screening according to any one of claims 1 to 3;
bone disease screening module: taking the color picture acquired by the ground reaction force image acquisition module as input, and calling a bone disease screening model to classify bone diseases;
the bone disease screening model is a machine learning model; the training process of the bone disease screening model comprises the following steps of:
obtaining a color picture by using the ground reaction force transformation method for disease screening according to any one of claims 1 to 3, and constructing a data set; and training the bone disease screening model by using the constructed data set to obtain a trained bone disease screening model.
6. The bone screening system of claim 5, wherein the bone screening model uses a Resnet classification model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310221888.0A CN115886787B (en) | 2023-03-09 | 2023-03-09 | Ground reaction force transformation method for disease screening, bone disease screening system and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310221888.0A CN115886787B (en) | 2023-03-09 | 2023-03-09 | Ground reaction force transformation method for disease screening, bone disease screening system and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115886787A CN115886787A (en) | 2023-04-04 |
| CN115886787B true CN115886787B (en) | 2023-06-02 |
Family
ID=86489987
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310221888.0A Active CN115886787B (en) | 2023-03-09 | 2023-03-09 | Ground reaction force transformation method for disease screening, bone disease screening system and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115886787B (en) |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6183425B1 (en) * | 1995-10-13 | 2001-02-06 | The United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration | Method and apparatus for monitoring of daily activity in terms of ground reaction forces |
| US10622102B2 (en) * | 2017-02-24 | 2020-04-14 | Siemens Healthcare Gmbh | Personalized assessment of bone health |
| KR102285315B1 (en) * | 2019-07-29 | 2021-08-04 | (주)아이유웰 | Foot pressure analyzer |
| CN114287915B (en) * | 2021-12-28 | 2024-03-05 | 深圳零动医疗科技有限公司 | Noninvasive scoliosis screening method and system based on back color images |
| CN115486839A (en) * | 2022-10-20 | 2022-12-20 | 南京市儿童医院 | Gait analysis and diagnosis system and method based on artificial intelligence |
| CN115553757B (en) * | 2022-12-06 | 2023-04-07 | 深圳市第二人民医院(深圳市转化医学研究院) | Method and equipment for intelligently screening effective ground reaction force in gait process |
-
2023
- 2023-03-09 CN CN202310221888.0A patent/CN115886787B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN115886787A (en) | 2023-04-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101740464B1 (en) | Method and system for diagnosis and prognosis of stroke and systme therefor | |
| TWI684997B (en) | Establishing method of bone age assessment and height prediction model, bone age assessment and height prediction system, and bone age assessment and height prediction method | |
| CN109770932A (en) | The processing method of multi-modal brain neuroblastoma image feature | |
| ES2960027T3 (en) | Detection of abnormalities in ECG signals | |
| CN110853111B (en) | Medical image processing system, model training method and training device | |
| CN113781440B (en) | Ultrasonic video focus detection method and device | |
| CN112001122B (en) | Non-contact physiological signal measurement method based on end-to-end generation countermeasure network | |
| CN113808106B (en) | Ultra-low dose PET image reconstruction system and method based on deep learning | |
| CN101909165B (en) | Mixed measurement-based video data panoramic imaging method | |
| CN107485383A (en) | A kind of speckle blood flow imaging method and apparatus based on constituent analysis | |
| CN111523617B (en) | Epilepsy detection system based on white matter fusion characteristic diagram and residual error attention network | |
| CN114241187A (en) | Muscle disease diagnosis system, device and medium based on ultrasonic bimodal images | |
| CN115886787B (en) | Ground reaction force transformation method for disease screening, bone disease screening system and equipment | |
| Lee et al. | Video analysis of human gait and posture to determine neurological disorders | |
| TWI663960B (en) | Foot deformity detection model, foot deformity detection system and foot deformity detection method | |
| CN111227834A (en) | Automatic rapid visualization method for resting brain function connection | |
| CN116934683A (en) | Method for assisting ultrasonic diagnosis of spleen wound by artificial intelligence | |
| Tang et al. | CNN-based automatic detection of bone conditions via diagnostic CT images for osteoporosis screening | |
| CN110827275A (en) | Liver nuclear magnetic artery phase image quality grading method based on raspberry group and deep learning | |
| CN115578419A (en) | Motion detection method and device, electronic equipment and storage medium | |
| KR102461893B1 (en) | Pet image medium-term data prediction apparatus based on least squares method for determining brain disease, and method thereof | |
| Senanayake et al. | Artificial neural network based gait patterns identification using neuromuscular signals and soft tissue deformation analysis of lower limbs muscles | |
| CN112155511B (en) | Method for compensating human eye shake in OCT acquisition process based on deep learning | |
| CN112837807A (en) | Early intelligent high-precision auxiliary diagnosis method for T2DM brain aging cognitive impairment | |
| Salehi et al. | Investigation and simulation of different medical image processing algorithms to improve image quality using simulink matlab |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |