CN113177359A - Dummy model-based body tissue state prediction method - Google Patents
Dummy model-based body tissue state prediction method Download PDFInfo
- Publication number
- CN113177359A CN113177359A CN202110481403.2A CN202110481403A CN113177359A CN 113177359 A CN113177359 A CN 113177359A CN 202110481403 A CN202110481403 A CN 202110481403A CN 113177359 A CN113177359 A CN 113177359A
- Authority
- CN
- China
- Prior art keywords
- dummy
- state
- muscle
- bilstm
- clustering
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention provides a body tissue state prediction method based on a dummy model, which comprises the following steps: under the condition of a limited work task, firstly dividing and determining action intentions; selecting from a plurality of preset ANYBODY dummy persons with the same posture but different speed settings according to the action intention; and in each marked intention time interval, keeping the motion path of the dummy consistent with the actually measured motion path, and gradually adjusting the overall action speed of the dummy until the muscle exertion prediction result and the muscle exertion state in the training set in the intention time interval reach the closest state. The muscle state prediction of the body tissue state prediction method based on the dummy model corresponds to the specific intention, and the precision is high.
Description
Technical Field
The invention relates to the technical field of human-computer interaction design and computer simulation, in particular to a human body tissue state prediction method based on a dummy model.
Background
The ANYBODY software provides a muscle and bone dummy which can accurately simulate two states of antagonism and relaxation, but the relationship between trunk muscles and external movement signals is unclear, and the situations that external behaviors are consistent and the internal tension degree is completely different exist.
The existing dummy model based on ANYBODY can simulate the human tissue state in the bending and stretching action by two modes: the finite element analysis of the first mode simulates a state dominated by antagonism, and the bending and stretching moment is borne by muscles, and the muscle activity is higher when the vertebral bend angle is larger. The simulation mode is that a single bending and stretching process of the dummy is divided into 16 discrete static state, muscle force is calculated under 16 static state respectively, each frame state is formed by respectively fitting the muscle and bone dummy, the force is calculated according to the static body balance state each time, and the electromyographic curve of each paraspinal muscle presents 'n' shape.
The finite element analysis of the second mode is to simulate the force application state of the human body in the continuous dynamic bending and stretching working process. In the process, the ANYBODY software can simulate the phenomenon of bending and relaxing, when trunk muscles bend and stretch, the muscles are forced to default to the state of bending and relaxing, and the trunk muscles are often switched into an antagonistic state due to fine operation intention, balance intention, external disturbance and the like. In the state of relaxing the waist bending, along with the increase of the body bending angle, the bending and stretching force moment is mainly balanced by passive tissues, the muscle force is reduced when the body bending angle is large, and the myoelectric curve of the paraspinal muscles which is dominant in the phenomenon of relaxing the waist bending approaches to a 'u' shape.
The simulation mode of the conventional ANYBODY software on the muscle control mode during bending and stretching is closer to the finite element analysis of the first mode when the simulation mode is closer to the static state, and the simulation mode is closer to the finite element analysis of the second mode when the simulation mode is more rapid to bend and stretch.
However, the finite element simulation algorithm provided by the conventional ANYBODY software is established based on a reverse dynamic balance formula, and the simulation result, whether an external force is added or not, is a reverse mechanical calculation result, and cannot identify whether the muscle control mode adopted by a human body at a certain moment is mainly antagonistic or relaxed, especially under the condition that the muscle control mode is influenced by a hand task and rapidly changes. The trunk state prediction and the tension degree identification firstly need to identify action intentions, and the intention identification cannot be provided by a finite element model.
Therefore, it is necessary to provide a dummy-model-based body tissue state prediction method that can recognize an action intention.
Disclosure of Invention
In view of the defects in the prior art, the invention aims to provide a human tissue state prediction method based on a dummy model, which has high precision and can identify action intentions.
In order to solve the problems, the technical scheme of the invention is as follows:
a body tissue state prediction method based on a dummy model comprises the following steps:
under the condition of a limited work task, firstly dividing and determining action intentions;
selecting from a plurality of preset ANYBODY dummy persons with the same posture but different speed settings according to the action intention; and
and in each marked intention time interval, keeping the motion path of the dummy consistent with the actually measured motion path, and gradually adjusting the overall action speed of the dummy until the muscle force generation prediction result and the muscle force generation state in the training set in the intention time interval reach the closest state.
Optionally, the step of keeping the motion path of the dummy consistent with the actually measured motion path in each labeled intention period, and gradually adjusting the overall motion speed of the dummy until the muscle exertion prediction result and the muscle exertion state in the training set in the intention period reach the closest state specifically includes:
taking the paraspinal muscles as input signals together, and outputting a one-dimensional two-step clustering result;
replacing the original electromyographic signals with cluster numbers, repeatedly clustering, and providing input vectors for the labeling step of the BilSTM-CRF by repeated clustering; and
and respectively correcting the predicted values of the muscle states in each clustering group number period according to the clustering group numbers output in the BiLSTM-CRF labeling step.
Optionally, in the step of replacing the original electromyographic signals with cluster numbers and repeatedly clustering, and providing input vectors for the BiLSTM-CRF labeling step by repeated clustering, the input vector construction method is as follows: the action label occurrence probability provided by BilSTM is used as the input of CRF, the input of each BilSTM unit is a corresponding single action vector in the behavior, and the BilSTM labels the action vector as a vector consisting of new labels by multiplying a weight matrix and adding an offset value.
Optionally, the step of respectively correcting the predicted values of the muscle states in each cluster group number period according to the cluster group numbers output in the BiLSTM-CRF labeling step specifically includes: and selecting the bending and stretching speed of the ANYBODY dummy according to the labeling result of the cluster number, wherein the speed is preset to 5 speed levels.
Optionally, the step of respectively correcting the predicted values of the muscle states in each cluster group number period according to the cluster group numbers output in the BiLSTM-CRF labeling step further includes: and in different intention periods, selecting one of the preset curves with the highest prediction correlation degree from the 5 different speed modes to replace the original prediction curve and correcting the prediction result in a segmented manner.
Compared with the prior art, the muscle state prediction of the body tissue state prediction method based on the dummy model is corresponding to the specific intention, the accuracy is much higher than that of the prediction method which does not correspond to the specific intention of the traditional finite element simulation method, and the condition that the form difference between the prediction and the actual measurement curve is obvious can not occur.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a block flow diagram of a method for predicting a condition of a body tissue based on a dummy model according to an embodiment of the present invention;
FIG. 2 is another block flow diagram of a method for predicting a condition of a body tissue based on a dummy model according to an embodiment of the present invention;
FIG. 3 is a comparison graph of different cluster subdivision results;
FIG. 4 is a block diagram of a process for repeating the clustering steps;
FIG. 5 is a comparison graph of measured and predicted right longissimus muscle;
FIG. 6 is a comparison graph of measured and predicted right trapezius muscle.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.
Fig. 1 is a block flow diagram of a body tissue state prediction method based on a dummy model according to an embodiment of the present invention, and as shown in fig. 1, the body tissue state prediction method based on the dummy model according to the present invention includes the following steps:
s1: under the condition of a limited work task, firstly dividing and determining action intentions;
s2: selecting from a plurality of preset ANYBODY dummy persons with the same posture but different speed settings according to the action intention; and
s3: and in each marked intention time interval, keeping the motion path of the dummy consistent with the actually measured motion path, and gradually adjusting the overall action speed of the dummy until the muscle force generation prediction result and the muscle force generation state in the training set in the intention time interval reach the closest state.
As shown in fig. 2, the step S3 specifically includes the following steps:
s21: taking the paraspinal muscles as input signals together, and outputting a one-dimensional two-step clustering result;
specifically, the minimum grouping number is set to be 3, rough classification is achieved, the minimum grouping number and the maximum grouping number of two-step clustering are increased, two-step clustering is repeated, multiple fine classification in different degrees is achieved, and multiple clustering results in the same time period are superposed. The sequence of subdivided cluster block numbers constitutes a coarse cluster number, this sequence being defined as a motion vector, the chronological arrangement of a set of motion vectors constituting a behavior.
The cluster number when the operation target is directly marked, started and ended and touched by hand is easy to be confirmed according to the motion signal and is used as a key cluster number, and the cluster number of the thick and thin clusters is rearranged, so that the cluster number and the average activation percentage of each tested muscle form a positive correlation relationship. In the two bending cycles, the pairs of different cluster subdivision results are shown in fig. 3.
S22: replacing the original electromyographic signals with cluster numbers and repeatedly clustering, wherein the repeated clustering provides input vectors for the step of labeling the BilSTM-CRF (Bi-directional Long Short Term Neural Network, Conditional Random Field);
the clustering method has the advantages that the clustering numbers are used for replacing original electromyographic signals and clustering is repeated, the problem of time axis difference is solved, muscle behaviors of different levels of different experimental objects can be compared with each other, the repeated clustering step is as shown in figure 4, trunk electromyographic data is collected, the electromyographic signals are converted into one-dimensional codes through preliminary clustering, when clustering results of different experimental objects are consistent, the grouping numbers of the preliminary clustering are increased, when clustering results of different experimental objects are inconsistent, the grouping numbers are increased, subdivision clustering is carried out, wherein the rough clustering represents commonality, and subdivision clustering reflects individual characteristics.
Each initial cluster number represents an action and is composed of a plurality of subdivided cluster block numbers. The motion vector composed by subdivision clustering coding is used as the input vector for the BilSTM-CRF layer.
The repeated clustering provides input vectors for the labeling step of the BilSTM-CRF, and the input vectors are constructed in the following way:
the word2vec toolkit is utilized to embed subdivision clustering codes into action vectors according to the maximum conditional probability, and the conditional probability relation among the arrangement signals can be constructed. In this model, the vector of each motion is inferred from the back-and-forth motion. The conditional probability of a specific action taking the combination of specific electromyographic signals as a known condition is shown as the following formula:
where P (action | EMG) represents the probability that a single action occurs under a given electromyographic signal, the occurrence of a single action means the occurrence of a series of cluster block number encodings. The key cluster number is used as a known condition, representing behaviors are obtained in advance, part of the key cluster number is obtained from a hand operation task setting, and tasks such as pinching, holding and accurate touching of hands can be confirmed in advance. The hand key cluster number and the cluster number representing the beginning and the end of stooping are used as the known key packet number of the Skip-gram algorithm, so that the embedding probability of all cluster numbers is calculated, and each cluster number is embedded into the action vector, thereby completing the construction of the action vector.
In addition, the Python3.7 and Tensor1.15 algorithms are adopted to complete sequence annotation based on a Bi-directional Long Short Term Neural Network (BilSTM) and a Random vector Field (CRF) model. In the data training step of the BilSTM labeling algorithm, the first two times of data in 3 identical bending and stretching tasks of each experimental object are taken as a training set, the last time is taken as a test set, and the training set is manually labeled, namely, each cluster number is labeled on the basis of the known key cluster number.
The action label occurrence probability provided by BilSTM is used as the input of CRF, the input of each BilSTM unit is a corresponding single action vector in the behavior, and the BilSTM labels the action vector as a vector consisting of new labels by multiplying a weight matrix and adding an offset value.
S23: and respectively correcting the predicted values of the muscle states in each clustering group number period according to the clustering group numbers output in the BiLSTM-CRF labeling step.
And selecting the bending and stretching speed of the total muscle and tendon dummy according to the labeling result of the cluster number, presetting the speed to 5 levels, and comparing the actually measured muscle curve after comparison and normalization (EMG is expressed in a percentage form) with predicted curves (also expressed in a percentage form) obtained by prediction at different speeds in the whole bending and stretching process time period. And when the Spearman correlation coefficient is obtained to be the highest, the preset value is considered to be used for predicting the states of tissues such as muscles in the intention time period, the force application mode is confirmed, and then the dummy model at the speed is selected to predict the stress states of all body tissues.
And in different intention periods, selecting one with the highest prediction correlation degree from the preset curves of 5 different force application modes to replace the original prediction curve, namely correcting the prediction result in a segmented manner. However, the difference of the absolute values of the small myoelectric signals measured actually is often large and related to the conductivity of the body surface medium at the electrode tip for myoelectric measurement, so that the range normalization process needs to be performed in an intentional period to make the range of the prediction consistent with the range of the measured curve.
In the following, taking the longest right muscle and the trapezius right muscle as examples, respectively, an RBF neural network is adopted, training sets are respectively selected according to each intention time interval divided by the step of BilSTM-CRF, and muscle activation degrees are predicted by taking motion signals such as spinal bend angles and the like as input.
Specifically, the result of prediction of the right longissimus muscle is shown in fig. 5, fig. 5 is a graph showing actual measurement and comparison of prediction of the right longissimus muscle, and as can be seen from fig. 5, the correlation between the predicted curve of the right longissimus muscle and the actual measured curve is high, and the effect of predicting the muscle state using the result of intention classification and recognition is also verified because the movement of the entire trunk has strong synchronism.
The result of prediction of the right trapezius muscle is shown in fig. 6, and fig. 6 is a graph showing actual measurement and comparison of prediction of the right trapezius muscle, and it can be seen from fig. 6 that the correlation between the predicted curve and the actual measurement curve of the right trapezius muscle is high, and in this embodiment, the correlation coefficient between the predicted curve and the actual measurement curve of each measured object is higher than 0.75.
The invention utilizes a bidirectional long-short time neural network (BILSTM) and a random vector field (CRF) labeling model, can complete human intention identification on the basis of a training set obtained by human body measurement based on electromyographic signal labeling action, and further adopts an RBF neural network to predict the electromyographic signal respectively based on a motion signal in each intention time period.
Compared with the prior art, the muscle state prediction of the body tissue state prediction method based on the dummy model corresponds to specific intentions, and the specific intentions can be obtained from trunk or hand behaviors, so that the accuracy of the method is much higher than that of the prediction method which does not correspond to the specific intentions of the traditional finite element simulation method, and the condition that the form difference between the prediction and the measured curve is obvious does not occur. In addition, the prediction result of the invention aims at all muscles, bones and ligaments of the ANYBODY dummy model, which cannot be realized by actual measurement.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (5)
1. A method for predicting a state of a body tissue based on a dummy model, the method comprising the steps of:
under the condition of a limited work task, firstly dividing and determining action intentions;
selecting from a plurality of preset ANYBODY dummy persons with the same posture but different speed settings according to the action intention; and
and in each marked intention time interval, keeping the motion path of the dummy consistent with the actually measured motion path, and gradually adjusting the overall action speed of the dummy until the muscle force generation prediction result and the muscle force generation state in the training set in the intention time interval reach the closest state.
2. The method for predicting the state of body tissues based on the dummy model as claimed in claim 1, wherein the step of keeping the motion path of the dummy consistent with the measured motion path and gradually adjusting the overall motion speed of the dummy until the predicted muscle force generation result reaches the closest state to the muscle force generation state in the training set in each marked intention period specifically comprises:
taking the paraspinal muscles as input signals together, and outputting a one-dimensional two-step clustering result;
replacing the original electromyographic signals with cluster numbers, repeatedly clustering, and providing input vectors for the labeling step of the BilSTM-CRF by repeated clustering; and
and respectively correcting the predicted values of the muscle states in each clustering group number period according to the clustering group numbers output in the BiLSTM-CRF labeling step.
3. The dummy model-based body tissue state prediction method according to claim 2, wherein the clustering number replaces the original electromyographic signal and the clustering is repeated, and in the step of providing the input vector for the BiLSTM-CRF labeling step, the input vector is constructed in a manner that: the action label occurrence probability provided by BilSTM is used as the input of CRF, the input of each BilSTM unit is a corresponding single action vector in the behavior, and the BilSTM labels the action vector as a vector consisting of new labels by multiplying a weight matrix and adding an offset value.
4. The method for predicting the state of body tissue based on a dummy model according to claim 2, wherein the step of correcting the predicted value of muscle state in each period of the cluster group number according to the cluster group number output in the labeling step of BilSTM-CRF specifically comprises: and selecting the bending and stretching speed of the ANYBODY dummy according to the labeling result of the cluster number, wherein the speed is preset to 5 speed levels.
5. The method for predicting the state of body tissue based on the dummy model according to claim 4, wherein the step of correcting the predicted value of muscle state in each period of the cluster group number according to the cluster group number output by the labeling step of BilSTM-CRF further comprises: and in different intention periods, selecting one of the preset curves with the highest prediction correlation degree from the 5 different speed modes to replace the original prediction curve and correcting the prediction result in a segmented manner.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110481403.2A CN113177359B (en) | 2021-04-30 | 2021-04-30 | Dummy model-based body tissue state prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110481403.2A CN113177359B (en) | 2021-04-30 | 2021-04-30 | Dummy model-based body tissue state prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113177359A true CN113177359A (en) | 2021-07-27 |
| CN113177359B CN113177359B (en) | 2023-04-18 |
Family
ID=76925927
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110481403.2A Active CN113177359B (en) | 2021-04-30 | 2021-04-30 | Dummy model-based body tissue state prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113177359B (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104382595A (en) * | 2014-10-27 | 2015-03-04 | 燕山大学 | Upper limb rehabilitation system and method based on myoelectric signal and virtual reality interaction technology |
| CN107397649A (en) * | 2017-08-10 | 2017-11-28 | 燕山大学 | A kind of upper limbs exoskeleton rehabilitation robot control method based on radial base neural net |
| US20180024634A1 (en) * | 2016-07-25 | 2018-01-25 | Patrick Kaifosh | Methods and apparatus for inferring user intent based on neuromuscular signals |
| CN108920893A (en) * | 2018-09-06 | 2018-11-30 | 南京医科大学 | A kind of cranio-maxillofacial bone and soft tissue form prediction method based on artificial intelligence |
| CN109394476A (en) * | 2018-12-06 | 2019-03-01 | 上海神添实业有限公司 | The automatic intention assessment of brain flesh information and upper limb intelligent control method and system |
| CN109480838A (en) * | 2018-10-18 | 2019-03-19 | 北京理工大学 | A kind of continuous compound movement Intention Anticipation method of human body based on surface layer electromyography signal |
| CN109998735A (en) * | 2019-01-25 | 2019-07-12 | 宁波创导三维医疗科技有限公司 | Silica gel nose augmentation prosthesis and its manufacturing method |
| US20190247650A1 (en) * | 2018-02-14 | 2019-08-15 | Bao Tran | Systems and methods for augmenting human muscle controls |
| CN110347837A (en) * | 2019-07-17 | 2019-10-18 | 电子科技大学 | A kind of unplanned Risk Forecast Method of being hospitalized again of cardiovascular disease |
| CN110400283A (en) * | 2018-04-20 | 2019-11-01 | 西门子医疗有限公司 | Real-time and accurate Soft Tissue Deformation prediction |
| CN111079418A (en) * | 2019-11-06 | 2020-04-28 | 科大讯飞股份有限公司 | Named body recognition method and device, electronic equipment and storage medium |
| AU2018344383A1 (en) * | 2017-10-02 | 2020-05-21 | Innosign B.V. | Assessment of Notch cellular signaling pathway activity using mathematical modelling of target gene expression |
| CA3123885A1 (en) * | 2018-12-20 | 2020-06-25 | Australian Institute of Robotic Orthopaedics Pty Ltd | Intelligent tissue classifier of bone and soft tissue |
| CN111547028A (en) * | 2020-04-20 | 2020-08-18 | 江苏大学 | Brake intensity fuzzy recognition method considering brake intention |
-
2021
- 2021-04-30 CN CN202110481403.2A patent/CN113177359B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104382595A (en) * | 2014-10-27 | 2015-03-04 | 燕山大学 | Upper limb rehabilitation system and method based on myoelectric signal and virtual reality interaction technology |
| US20180024634A1 (en) * | 2016-07-25 | 2018-01-25 | Patrick Kaifosh | Methods and apparatus for inferring user intent based on neuromuscular signals |
| CN107397649A (en) * | 2017-08-10 | 2017-11-28 | 燕山大学 | A kind of upper limbs exoskeleton rehabilitation robot control method based on radial base neural net |
| AU2018344383A1 (en) * | 2017-10-02 | 2020-05-21 | Innosign B.V. | Assessment of Notch cellular signaling pathway activity using mathematical modelling of target gene expression |
| US20190247650A1 (en) * | 2018-02-14 | 2019-08-15 | Bao Tran | Systems and methods for augmenting human muscle controls |
| CN110400283A (en) * | 2018-04-20 | 2019-11-01 | 西门子医疗有限公司 | Real-time and accurate Soft Tissue Deformation prediction |
| CN108920893A (en) * | 2018-09-06 | 2018-11-30 | 南京医科大学 | A kind of cranio-maxillofacial bone and soft tissue form prediction method based on artificial intelligence |
| CN109480838A (en) * | 2018-10-18 | 2019-03-19 | 北京理工大学 | A kind of continuous compound movement Intention Anticipation method of human body based on surface layer electromyography signal |
| CN109394476A (en) * | 2018-12-06 | 2019-03-01 | 上海神添实业有限公司 | The automatic intention assessment of brain flesh information and upper limb intelligent control method and system |
| CA3123885A1 (en) * | 2018-12-20 | 2020-06-25 | Australian Institute of Robotic Orthopaedics Pty Ltd | Intelligent tissue classifier of bone and soft tissue |
| CN109998735A (en) * | 2019-01-25 | 2019-07-12 | 宁波创导三维医疗科技有限公司 | Silica gel nose augmentation prosthesis and its manufacturing method |
| CN110347837A (en) * | 2019-07-17 | 2019-10-18 | 电子科技大学 | A kind of unplanned Risk Forecast Method of being hospitalized again of cardiovascular disease |
| CN111079418A (en) * | 2019-11-06 | 2020-04-28 | 科大讯飞股份有限公司 | Named body recognition method and device, electronic equipment and storage medium |
| CN111547028A (en) * | 2020-04-20 | 2020-08-18 | 江苏大学 | Brake intensity fuzzy recognition method considering brake intention |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113177359B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tsinganos et al. | Improved gesture recognition based on sEMG signals and TCN | |
| Rane et al. | Deep learning for musculoskeletal force prediction | |
| CN110710970B (en) | Method and device for recognizing limb actions, computer equipment and storage medium | |
| CN112861604A (en) | Myoelectric action recognition and control method irrelevant to user | |
| CN110348055B (en) | Method for obtaining and optimizing material parameters of Chaboche viscoplasticity constitutive model | |
| Mochurad et al. | Modeling of psychomotor reactions of a person based on modification of the tapping test | |
| WO2021142532A1 (en) | Activity recognition with deep embeddings | |
| CN113177359B (en) | Dummy model-based body tissue state prediction method | |
| Terpstra | On estimating a population proportion via ranked set sampling | |
| CN117034123B (en) | Fault monitoring system and method for fitness equipment | |
| Rahagiyanto et al. | Hand gesture classification for sign language using artificial neural network | |
| CN113116363A (en) | Method for judging hand fatigue degree based on surface electromyographic signals | |
| CN109948465A (en) | A kind of surface electromyogram signal classification method based on multiple target | |
| Lee et al. | A neuro-fuzzy model for estimating electromyographical activity of trunk muscles due to manual lifting | |
| CN110598789A (en) | Human fatigue state prediction method and system based on fuzzy perceptron | |
| CN116138794A (en) | Low training burden myoelectric mode identification method and system based on impulse neural network | |
| CN115981461A (en) | Electromyography control method based on multitask learning Transformer | |
| Guo et al. | A novel fuzzy neural network-based rehabilitation stage classifying method for the upper limb rehabilitation robotic system | |
| Zhao et al. | Extreme Gradient Boosting for Surface Electromyography Classification on Time-Domain Features | |
| CN112764524A (en) | Myoelectric signal gesture action recognition method based on texture features | |
| Dobrea et al. | An EEG coherence based method used for mental tasks classification | |
| Li et al. | sEMG and IMU Data-based Hand Gesture Recognition Method using Multi-stream CNN with a Fine-tuning Transfer Framework | |
| CN113040710B (en) | Sleep state analysis method and device | |
| Zhu et al. | Sentiment recognition model of EEG signals combined with one-dimensional convolution and BiBASRU-AT | |
| West et al. | Complete imputation of missing repeated categorical data: one‐sample applications |
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 |