CN111544005A - Parkinson's disease dyskinesia quantification and identification method based on support vector machine - Google Patents
Parkinson's disease dyskinesia quantification and identification method based on support vector machine Download PDFInfo
- Publication number
- CN111544005A CN111544005A CN202010414448.3A CN202010414448A CN111544005A CN 111544005 A CN111544005 A CN 111544005A CN 202010414448 A CN202010414448 A CN 202010414448A CN 111544005 A CN111544005 A CN 111544005A
- Authority
- CN
- China
- Prior art keywords
- signal
- parkinson
- support vector
- vector machine
- tested object
- 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
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012706 support-vector machine Methods 0.000 title claims abstract description 46
- 238000011002 quantification Methods 0.000 title claims abstract description 23
- 208000012661 Dyskinesia Diseases 0.000 title claims description 31
- 208000018737 Parkinson disease Diseases 0.000 title claims description 25
- 230000005021 gait Effects 0.000 claims abstract description 63
- 210000003423 ankle Anatomy 0.000 claims abstract description 41
- 210000000707 wrist Anatomy 0.000 claims abstract description 41
- 238000012952 Resampling Methods 0.000 claims abstract description 6
- 230000002194 synthesizing effect Effects 0.000 claims abstract description 5
- 239000013598 vector Substances 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000001133 acceleration Effects 0.000 claims description 7
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 238000003786 synthesis reaction Methods 0.000 claims description 6
- 238000000546 chi-square test Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 2
- 208000016285 Movement disease Diseases 0.000 abstract description 13
- 230000006870 function Effects 0.000 description 11
- 238000003745 diagnosis Methods 0.000 description 6
- 208000027089 Parkinsonian disease Diseases 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 206010034010 Parkinsonism Diseases 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 208000024891 symptom Diseases 0.000 description 3
- 206010044565 Tremor Diseases 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 206010006100 Bradykinesia Diseases 0.000 description 1
- 208000027776 Extrapyramidal disease Diseases 0.000 description 1
- 208000006083 Hypokinesia Diseases 0.000 description 1
- 208000002740 Muscle Rigidity Diseases 0.000 description 1
- 208000012902 Nervous system disease Diseases 0.000 description 1
- 208000025966 Neurological disease Diseases 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 210000000544 articulatio talocruralis Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001684 chronic effect Effects 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 210000002683 foot Anatomy 0.000 description 1
- 210000003141 lower extremity Anatomy 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000037230 mobility Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 210000001364 upper extremity Anatomy 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/112—Gait analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1118—Determining activity level
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1121—Determining geometric values, e.g. centre of rotation or angular range of movement
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
- A61B5/4082—Diagnosing or monitoring movement diseases, e.g. Parkinson, Huntington or Tourette
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6813—Specially adapted to be attached to a specific body part
- A61B5/6824—Arm or wrist
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6813—Specially adapted to be attached to a specific body part
- A61B5/6829—Foot or ankle
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Physiology (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Neurology (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Neurosurgery (AREA)
- Developmental Disabilities (AREA)
- Geometry (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention belongs to the technical field of signal identification, and particularly relates to a parkinsonian movement disorder quantification and identification method based on a support vector machine, aiming at solving the problem that the prior art cannot accurately realize the parkinsonian movement disorder quantification and identification. The invention comprises the following steps: acquiring wrist movement signals and ankle movement signals of a tested object and a healthy person; resampling and synthesizing the signals; extracting walking intervals through a sliding window; respectively extracting gait characteristics of a tested object and a healthy person; normalizing gait characteristics and classifying the gait characteristics through a trained support vector machine; and calculating Pr values, wherein the intervals Pr of the interval Pr of more than 0.9, 0.9 of more than or equal to 0.6, 0.6 of more than or equal to 0.5 and Pr of less than 0.5 respectively correspond to the tested objects of severe Parkinson patients, moderate Parkinson patients, mild Parkinson patients and non-Parkinson patients. The parkinsonian movement disorder quantitative and identification method has the advantages of high accuracy, high precision and less occupied resources, is also suitable for remote medical treatment, reduces the cost and improves the efficiency.
Description
Technical Field
The invention belongs to the technical field of signal identification, and particularly relates to a Parkinson's disease dyskinesia quantification and identification method based on a support vector machine.
Background
Parkinson's disease is a chronic neurological disease characterized primarily by bradykinesia, tremor, rigidity, and abnormalities in posture and gait. Currently, there are over 600 million people with Parkinson's disease worldwide, most of which are elderly, who require much longer detection times than younger people. In some medical researches, the extrapyramidal diseases represented by the Parkinson's disease have the characteristics of high morbidity, high disability rate, complex and various dyskinesia forms and the like. In clinical work, the severity and judgment of a Parkinson's disease patient by a clinician depends mainly on clinical experience, but it is difficult to provide objective and reliable quantitative evaluation because the clinical experience is too subjective.
In the traditional research, the quantitative analysis method for the movement of the Parkinson patients mainly depends on footprint analysis, and although the method is simple and practical and is not limited by test site conditions, the measurement precision is low, and the method is seriously dependent on the analysis of professionals, so that the requirement of modern movement evaluation cannot be met.
Generally, no perfect solution exists for the quantification method of the parkinsonian movement disorder at present, and a method capable of accurately quantifying and identifying the parkinsonian movement disorder is urgently needed in the field.
Disclosure of Invention
In order to solve the above problems in the prior art, that is, the prior art cannot accurately realize the quantification and identification of the parkinsonian movement disorder, the invention provides a method for quantifying and identifying the parkinsonian movement disorder based on a support vector machine, which comprises the following steps:
step S10, acquiring wrist movement signals and ankle movement signals of the tested object and the healthy person when walking, standing and turning respectively through inertial sensors arranged on the wrist and the ankle of the tested object and the healthy person;
step S20, resampling the wrist motion signal and the ankle motion signal, and synthesizing the resampled signals into an accelerometer signal by a preset first algorithm;
step S30, acquiring a walking interval through a preset first sliding window based on the signal of the accelerometer signal Z axis;
step S40, based on the re-sampled wrist movement signal and ankle movement signal, respectively extracting the gait characteristics of the tested object and the healthy person in the walking interval through a preset second sliding window;
step S50, the gait features are normalized into gait feature vectors, classification of the gait feature vectors is carried out through a trained support vector machine, and the positive rate Pr of the tested object relative to the healthy person is calculated through a preset second algorithm;
step S60, if Pr is more than 0.9, the tested object is a severe Parkinson patient; if 0.9 & gtPr is more than or equal to 0.6, the tested object is a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, the tested object is a mild Parkinson patient; if Pr is less than 0.5, the subject is a non-Parkinson patient.
In some preferred embodiments, the preset first algorithm is:
wherein As is the resultant accelerometer signal, Ax、Ay、AzRespectively, representing the acceleration signals of the axial components of the three-axis accelerometer X, Y, Z in the inertial sensor.
In some preferred embodiments, step S30 is preceded by a step of signal filtering, which includes:
and filtering the accelerometer signal, the re-sampled wrist motion signal and the ankle motion signal by a fifth-order elliptical band-pass filter.
In some preferred embodiments, step S30 includes:
step S31, calculating the signal power of the Z axis of the accelerometer signal through a preset first sliding window, and taking a point corresponding to the signal power larger than a set upper limit as a starting position and a point corresponding to the signal power smaller than a set lower limit as an ending position;
and step S32, calculating the distance between the starting position and the ending position, and taking the distance larger than a set threshold value as a time interval to obtain a walking interval.
In some preferred embodiments, step S40 includes:
step S41, respectively acquiring window data of the wrist movement signal and the ankle movement signal of the tested object and the healthy person, which are re-sampled, in the walking interval through a preset second sliding window;
and step S42, taking the window data as a data frame, and respectively extracting the gait characteristics of the tested object and the healthy person in the data frame.
In some preferred embodiments, step S50 is preceded by a step of gait feature screening, which includes:
and respectively selecting the gait characteristics of the tested object and the set number of healthy people as the optimal gait characteristics by a chi-square test method.
In some preferred embodiments, the kernel function of the support vector machine is a radial basis function, and the radial basis function is:
wherein x isi、xjRespectively represent gait feature vectors of a Parkinson patient and a healthy person,
is the bandwidth of the radial basis function greater than 0.
In some preferred embodiments, the preset second algorithm is:
wherein D ispThe number of positive results output for the SVM classifier, D represents the data acquired in the second sliding window within the set time periodThe total number of frames.
On the other hand, the invention provides a parkinsonian movement disorder quantification and identification system based on a support vector machine, which comprises a data acquisition module, a signal synthesis module, a data distinguishing module, a gait feature extraction module, a classification and positive rate calculation module and an output module;
the data acquisition module is configured to acquire wrist movement signals and ankle movement signals of the tested object and the healthy person during walking, standing and turning respectively through inertial sensors arranged on the wrist and the ankle of the tested object and the healthy person;
the signal synthesis module is configured to resample the wrist movement signal and the ankle movement signal and synthesize the resampled signals into an accelerometer signal through a preset first algorithm;
the data distinguishing module is configured to obtain a walking interval through a preset first sliding window based on a signal of the accelerometer signal Z axis;
the gait feature extraction module is configured to extract gait features of the measured object and the healthy person in the walking interval through a preset second sliding window based on the re-sampled wrist motion signal and the re-sampled ankle motion signal;
the classification and positive rate calculation module is configured to normalize the gait features into gait feature vectors, classify the gait feature vectors through a trained support vector machine, and calculate the positive rate Pr of the tested object relative to the healthy person through a preset second algorithm;
the output module is configured to output that the detected object is a severe Parkinson patient if Pr is greater than 0.9; if 0.9 & gtPr is more than or equal to 0.6, outputting the tested object as a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, outputting the tested object as a mild Parkinson patient; if Pr is less than 0.5, the output object is the non-Parkinson patient.
In a third aspect of the present invention, a storage device is provided, in which a plurality of programs are stored, the programs being suitable for being loaded and executed by a processor to implement the above-mentioned method for quantifying and identifying dyskinesia of parkinson's disease based on a support vector machine.
In a fourth aspect of the present invention, a processing apparatus is provided, which includes a processor, a storage device; the processor is suitable for executing various programs; the storage device is suitable for storing a plurality of programs; the program is suitable to be loaded and executed by a processor to realize the parkinsonian movement disorder quantification and identification method based on the support vector machine.
The invention has the beneficial effects that:
(1) the parkinsonism dyskinesia quantification and identification method based on the support vector machine fully utilizes the excellent performance of the support vector machine in the nonlinear problem classification of small sample data, reduces the calculation complexity by resampling the acquired signals, eliminates the influence of the gravity component and high-frequency noise of the signals by the filter, fully combines the characteristics of the parkinsonism person, screens the optimal gait characteristics by the chi-square test method, improves the accuracy and precision of parkinsonism dyskinesia quantification and identification, and reduces the occupation of system resources.
(2) The parkinsonian movement disorder quantification and identification method based on the support vector machine is also suitable for remote medical diagnosis, a patient can carry out wrist movement signals and ankle movement signals by himself or by the assistance of other people, and the parkinsonian movement disorder quantification and identification are carried out by the method, so that a new solution direction is provided for the diagnosis of the parkinsonian diseases, the medical cost is reduced, and the diagnosis efficiency is improved.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:
FIG. 1 is a flow chart diagram of the Parkinson's disease dyskinesia quantification and identification method based on a support vector machine.
Detailed Description
The present application will be described in further detail with reference to the following drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
The invention relates to a Parkinson's disease people dyskinesia quantification and identification method based on a support vector machine, which comprises the following steps:
step S10, acquiring wrist movement signals and ankle movement signals of the tested object and the healthy person when walking, standing and turning respectively through inertial sensors arranged on the wrist and the ankle of the tested object and the healthy person;
step S20, resampling the wrist motion signal and the ankle motion signal, and synthesizing the resampled signals into an accelerometer signal by a preset first algorithm;
step S30, acquiring a walking interval through a preset first sliding window based on the signal of the accelerometer signal Z axis;
step S40, based on the re-sampled wrist movement signal and ankle movement signal, respectively extracting the gait characteristics of the tested object and the healthy person in the walking interval through a preset second sliding window;
step S50, the gait features are normalized into gait feature vectors, classification of the gait feature vectors is carried out through a trained support vector machine, and the positive rate Pr of the tested object relative to the healthy person is calculated through a preset second algorithm;
step S60, if Pr is more than 0.9, the tested object is a severe Parkinson patient; if 0.9 & gtPr is more than or equal to 0.6, the tested object is a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, the tested object is a mild Parkinson patient; if Pr is less than 0.5, the subject is a non-Parkinson patient.
In order to more clearly describe the method for quantifying and identifying the dyskinesia of the parkinson's disease patient based on the support vector machine, the following describes the steps in the embodiment of the invention in detail with reference to fig. 1.
The Parkinson' S disease dyskinesia quantification and identification method based on the support vector machine comprises the steps of S10-S60, and the steps are described in detail as follows:
in step S10, the inertial sensors provided on the wrist and the ankle of the subject and the healthy person respectively acquire a wrist movement signal and an ankle movement signal when the subject and the healthy person walk, stand, and turn around.
In one embodiment of the present invention, wrist and ankle motion signals were collected for 18 Parkinson's and 10 healthy individuals during walking, standing and turning.
The motion signals are acquired by inertial sensors arranged on the wrist and ankle of the object to be tested (Parkinson's disease person and healthy person).
In addition, the other common symptom of the Parkinson patients is dyskinesia, and is characterized in that the dyskinesia is slow in movement, the dyskinesia is difficult to start moving, the mobility is lost and the like, in clinic, the dyskinesia of gait is characterized in that the dyskinesia symptom is slow in movement, the dyskinesia of gait is usually accompanied by walking slowly and dragging the feet frequently, the swing amplitude of the arms is gradually reduced or even disappears, and the stride is reduced. Therefore, the motion signals of the wrist and the ankle are selected as the basis for quantifying and identifying the dyskinesia of the Parkinson patients.
The features of standard deviation, absolute value of energy, maximum value and quantile (0.9) on the inertial sensor of the hand reflect the upper limb swing of the subject during walking, the number of crossing zero-crossing points reflects the symptoms of tremor, and the quantile (0.9) and maximum value features of the inertial sensor on the ankle joint reflect the movement ability of the lower limb.
And step S20, resampling the wrist motion signal and the ankle motion signal, and synthesizing the resampled signals into an accelerometer signal by a preset first algorithm.
In one embodiment of the invention, the collected wrist motion signal and ankle motion signal of the parkinsonian and healthy people during walking, standing and turning are respectively resampled to 50Hz to reduce the calculation complexity, and the resampled signals are synthesized into an accelerometer signal by a preset first algorithm, as shown in formula (1):
wherein As is the resultant accelerometer signal, Ax、Ay、AzRespectively, representing the acceleration signals of the axial components of the three-axis accelerometer X, Y, Z in the inertial sensor.
Before the walking interval acquisition in step S30, the accelerometer signal, the resampled wrist motion signal, and the ankle motion signal may be filtered by a fifth-order elliptic band pass filter to eliminate the influence of the gravity component and the high frequency noise.
And step S30, acquiring a walking interval through a preset first sliding window based on the signal of the accelerometer signal Z axis.
Step S31, calculating the signal power of the Z-axis of the accelerometer signal through a preset first sliding window, and taking a point corresponding to the signal power greater than a set upper limit as a start position and a point corresponding to the signal power less than a set lower limit as an end position. In an embodiment of the present invention, the length of the preset first sliding window is set to 20 sampling points.
And step S32, calculating the distance between the starting position and the ending position, and taking the distance larger than a set threshold value as a time interval to obtain a walking interval. In one embodiment of the present invention, if the distance between the start position and the end position is greater than 256 sampling points (i.e. the walking time is greater than 5s), a time interval is divided.
And step S40, respectively extracting the gait characteristics of the tested object and the healthy person in the walking interval through a preset second sliding window based on the re-sampled wrist movement signal and ankle movement signal.
And step S41, acquiring window data of the wrist movement signal and the ankle movement signal of the tested object and the healthy person respectively in the walking interval through a preset second sliding window.
In an embodiment of the present invention, the preset size of the second sliding window is 256, the sliding step is 16 sampling points, and the window data of the motion signal captured by this sliding window is processed into a data frame.
And step S42, taking the window data as a data frame, and respectively extracting the gait characteristics of the tested object and the healthy person in the data frame.
In order to extract the optimal gait characteristics, before characteristic normalization, gait characteristics of a tested object and a healthy person are respectively screened by a chi-square test method, and finally 8 gait characteristics of different tested objects and healthy persons are obtained, wherein the gait characteristics are respectively the data standard deviation of a Z axis of a gyroscope on a wrist inertial sensor, the energy absolute value of the Z axis of the gyroscope, the zero crossing point number of a Y axis of the gyroscope, the extreme value of a synthetic acceleration signal, 0.9 quantile of the synthetic acceleration signal, the extreme value of a synthetic acceleration signal of an ankle and 0.9 quantile of the synthetic acceleration signal.
Step S50, the gait features are normalized into gait feature vectors, classification of the gait feature vectors is carried out through a trained support vector machine, and the positive rate Pr of the tested object relative to the healthy person is calculated through a preset second algorithm.
Normalizing the gait features into gait feature vectors, as shown in equation (2):
wherein, XscaledRepresenting the normalized gait feature vector, X representing the gait feature to be normalized, Xmax、XminRespectively a maximum value and a minimum value in the gait characteristics to be normalized.
The data segments of the Parkinson patients and the healthy people are classified by using a support vector machine, the support vector machine is suitable for solving the problem of nonlinear classification of small samples, a nonlinear relation is presented between class labels and extracted features, so a Radial Basis Function (RFB) is used as a kernel Function of a support vector machine classifier, the RBF kernel Function has the capability of approximating a nonlinear Function, can process an insolubility rule in a system, and has good generalization capability and fast learning speed, as shown in formula (3):
wherein x isi、xjRespectively represent gait feature vectors of a Parkinson patient and a healthy person,
is the bandwidth of the radial basis function greater than 0.
Classifying the gait feature vectors through a trained support vector machine, and calculating the positive rate Pr of the tested object relative to the healthy person as shown in formula (4)
Wherein D ispD represents the total number of data frames acquired by the second sliding window within the set time period.
Pr represents the ratio of scan data segments considered to belong to parkinson's patients by the support vector machine classifier within a 30 second (1500 data points) data length of the SVM output.
Step S60, if Pr is more than 0.9, the tested object is a severe Parkinson patient; if 0.9 & gtPr is more than or equal to 0.6, the tested object is a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, the tested object is a mild Parkinson patient; if Pr is less than 0.5, the subject is a non-Parkinson patient.
The invention also evaluates the performance of the support vector machine classifier by a ten-fold cross validation method, and adopts the following common performance indexes: accuracy, recall, F1 score. The parkinson patients were set as positive group and the healthy persons as negative group, wherein accuracy was the ratio of correctly classifying the parkinson patients and the healthy persons, recall was the ratio of the true positive group, and F1 score was the weighted average of model precision and recall. The present invention calculates that the classification accuracy of parkinson's patients to healthy subjects is about 92.9%, the sensitivity (the ratio of classifying positive examples into positive examples by the classifier, i.e., the ratio of classifying parkinson's patients into patients) is 88.9%, and the specificity (the ratio of classifying negative examples into negative examples, i.e., the ratio of classifying healthy patients into healthy patients) is 100%, showing the good ability of the present invention method in identifying parkinson's patients from healthy patients.
And finally, carrying out quantitative reliability evaluation on the method through a secondary weighting Kappa coefficient, wherein the reliability evaluation is shown as a formula (5):
wherein, ∑ ω f0s is the sum of the observed weighted frequencies in the ordinal scale, ∑ ω fcIs the sum of randomly expected weighted frequencies in the ordinal scale.
The calculation method of the quadratic weight ω is shown in formula (6):
where i-j is the number of manual classifications that are inconsistent with the method of the present invention, and k is the number of levels on the ordinal scale.
In an embodiment of the present invention, k is 4, and ω on 4 levels is 1, 0.89, 0.56, and 0, respectively, which indicates that the larger the evaluation deviation is, the larger the weight is. Through experimental tests, compared with expert diagnosis, the secondary weighting Kappa coefficient k of the method is 0.961, and the method has higher consistency with expert manual diagnosis and high accuracy of diagnosis results.
The parkinsonian movement disorder quantification and identification system based on the support vector machine comprises a data acquisition module, a signal synthesis module, a data distinguishing module, a gait feature extraction module, a classification and positive rate calculation module and an output module;
the data acquisition module is configured to acquire wrist movement signals and ankle movement signals of the tested object and the healthy person during walking, standing and turning respectively through inertial sensors arranged on the wrist and the ankle of the tested object and the healthy person;
the signal synthesis module is configured to resample the wrist movement signal and the ankle movement signal and synthesize the resampled signals into an accelerometer signal through a preset first algorithm;
the data distinguishing module is configured to obtain a walking interval through a preset first sliding window based on a signal of the accelerometer signal Z axis;
the gait feature extraction module is configured to extract gait features of the measured object and the healthy person in the walking interval through a preset second sliding window based on the re-sampled wrist motion signal and the re-sampled ankle motion signal;
the classification and positive rate calculation module is configured to normalize the gait features into gait feature vectors, classify the gait feature vectors through a trained support vector machine, and calculate the positive rate Pr of the tested object relative to the healthy person through a preset second algorithm;
the output module is configured to output that the detected object is a severe Parkinson patient if Pr is greater than 0.9; if 0.9 & gtPr is more than or equal to 0.6, outputting the tested object as a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, outputting the tested object as a mild Parkinson patient; if Pr is less than 0.5, the output object is the non-Parkinson patient.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working process and related description of the system described above may refer to the corresponding process in the foregoing method embodiments, and will not be described herein again.
It should be noted that, the parkinson's disease dyskinesia quantification and identification system based on the support vector machine provided in the above embodiment is only illustrated by the division of the above functional modules, and in practical applications, the above functions may be allocated to different functional modules according to needs, that is, the modules or steps in the embodiment of the present invention are further decomposed or combined, for example, the modules in the above embodiment may be combined into one module, or may be further split into a plurality of sub-modules, so as to complete all or part of the above described functions. The names of the modules and steps involved in the embodiments of the present invention are only for distinguishing the modules or steps, and are not to be construed as unduly limiting the present invention.
A storage device according to a third embodiment of the present invention stores a plurality of programs, which are suitable for being loaded and executed by a processor to realize the above-mentioned method for quantifying and identifying dyskinesia of parkinson's disease based on a support vector machine.
A processing apparatus according to a fourth embodiment of the present invention includes a processor, a storage device; a processor adapted to execute various programs; a storage device adapted to store a plurality of programs; the program is suitable to be loaded and executed by a processor to realize the parkinsonian movement disorder quantification and identification method based on the support vector machine.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes and related descriptions of the storage device and the processing device described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
Those of skill in the art would appreciate that the various illustrative modules, method steps, and modules described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that programs corresponding to the software modules, method steps may be located in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. To clearly illustrate this interchangeability of electronic hardware and software, various illustrative components and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as electronic hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The terms "first," "second," and the like are used for distinguishing between similar elements and not necessarily for describing or implying a particular order or sequence.
The terms "comprises," "comprising," or any other similar term are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can fall into the protection scope of the invention.
Claims (11)
1. A Parkinson's disease dyskinesia quantification and identification method based on a support vector machine is characterized by comprising the following steps:
step S10, acquiring wrist movement signals and ankle movement signals of the tested object and the healthy person when walking, standing and turning respectively through inertial sensors arranged on the wrist and the ankle of the tested object and the healthy person;
step S20, resampling the wrist motion signal and the ankle motion signal, and synthesizing the resampled signals into an accelerometer signal by a preset first algorithm;
step S30, acquiring a walking interval through a preset first sliding window based on the signal of the accelerometer signal Z axis;
step S40, based on the re-sampled wrist movement signal and ankle movement signal, respectively extracting the gait characteristics of the tested object and the healthy person in the walking interval through a preset second sliding window;
step S50, the gait features are normalized into gait feature vectors, classification of the gait feature vectors is carried out through a trained support vector machine, and the positive rate Pr of the tested object relative to the healthy person is calculated through a preset second algorithm;
step S60, if Pr is more than 0.9, the tested object is a severe Parkinson patient; if 0.9 & gtPr is more than or equal to 0.6, the tested object is a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, the tested object is a mild Parkinson patient; if Pr is less than 0.5, the subject is a non-Parkinson patient.
2. The method for quantifying and identifying dyskinesia of Parkinson's disease patients based on support vector machine according to claim 1, wherein the preset first algorithm is as follows:
wherein As is the resultant accelerometer signal, Ax、Ay、AzRespectively, representing the acceleration signals of the axial components of the three-axis accelerometer X, Y, Z in the inertial sensor.
3. The method for quantifying and identifying dyskinesia of Parkinson' S disease based on support vector machine according to claim 1, wherein step S30 is preceded by a step of signal filtering, which comprises:
and filtering the accelerometer signal, the re-sampled wrist motion signal and the ankle motion signal by a fifth-order elliptical band-pass filter.
4. The method for quantifying and identifying dyskinesia of Parkinson' S disease based on support vector machine according to claim 1, wherein step S30 comprises:
step S31, calculating the signal power of the Z axis of the accelerometer signal through a preset first sliding window, and taking a point corresponding to the signal power larger than a set upper limit as a starting position and a point corresponding to the signal power smaller than a set lower limit as an ending position;
and step S32, calculating the distance between the starting position and the ending position, and taking the distance larger than a set threshold value as a time interval to obtain a walking interval.
5. The method for quantifying and identifying dyskinesia of Parkinson' S disease based on support vector machine according to claim 1, wherein step S40 comprises:
step S41, respectively acquiring window data of the wrist movement signal and the ankle movement signal of the tested object and the healthy person, which are re-sampled, in the walking interval through a preset second sliding window;
and step S42, taking the window data as a data frame, and respectively extracting the gait characteristics of the tested object and the healthy person in the data frame.
6. The method for quantifying and identifying dyskinesia of Parkinson' S disease patients based on support vector machine according to claim 1, wherein step S50 is preceded by a step of gait feature screening, the method comprises:
and respectively selecting the gait characteristics of the tested object and the set number of healthy people as the optimal gait characteristics by a chi-square test method.
7. The method for quantifying and identifying dyskinesia of Parkinson's disease based on support vector machine according to claim 1, wherein the kernel function of the support vector machine is radial basis function, and the radial basis function is:
wherein x isi、xjRespectively represent gait feature vectors of a Parkinson patient and a healthy person,
is the bandwidth of the radial basis function greater than 0.
8. The method for quantifying and identifying dyskinesia of Parkinson's disease patients based on support vector machine according to claim 1, wherein the preset second algorithm is as follows:
wherein D ispD represents the total number of data frames acquired by the second sliding window within the set time period.
9. A parkinsonian dyskinesia quantification and identification system based on a support vector machine is characterized by comprising a data acquisition module, a signal synthesis module, a data distinguishing module, a gait feature extraction module, a classification and positive rate calculation module and an output module;
the data acquisition module is configured to acquire wrist movement signals and ankle movement signals of the tested object and the healthy person during walking, standing and turning respectively through inertial sensors arranged on the wrist and the ankle of the tested object and the healthy person;
the signal synthesis module is configured to resample the wrist movement signal and the ankle movement signal and synthesize the resampled signals into an accelerometer signal through a preset first algorithm;
the data distinguishing module is configured to obtain a walking interval through a preset first sliding window based on a signal of the accelerometer signal Z axis;
the gait feature extraction module is configured to extract gait features of the measured object and the healthy person in the walking interval through a preset second sliding window based on the re-sampled wrist motion signal and the re-sampled ankle motion signal;
the classification and positive rate calculation module is configured to normalize the gait features into gait feature vectors, classify the gait feature vectors through a trained support vector machine, and calculate the positive rate Pr of the tested object relative to the healthy person through a preset second algorithm;
the output module is configured to output that the detected object is a severe Parkinson patient if Pr is greater than 0.9; if 0.9 & gtPr is more than or equal to 0.6, outputting the tested object as a moderate Parkinson patient; if 0.6 & gt Pr is more than or equal to 0.5, outputting the tested object as a mild Parkinson patient; if Pr is less than 0.5, the output object is the non-Parkinson patient.
10. A storage means having stored therein a plurality of programs, characterized in that said programs are adapted to be loaded and executed by a processor to implement the support vector machine based parkinson's disease dyskinesia quantification and identification method of any of claims 1-8.
11. A treatment apparatus comprises
A processor adapted to execute various programs; and
a storage device adapted to store a plurality of programs;
wherein the program is adapted to be loaded and executed by a processor to perform:
the method for quantification and identification of dyskinesia of Parkinson's disease based on support vector machine according to any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010414448.3A CN111544005B (en) | 2020-05-15 | 2020-05-15 | Parkinson's disease dyskinesia quantification and identification method based on support vector machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010414448.3A CN111544005B (en) | 2020-05-15 | 2020-05-15 | Parkinson's disease dyskinesia quantification and identification method based on support vector machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111544005A true CN111544005A (en) | 2020-08-18 |
| CN111544005B CN111544005B (en) | 2022-03-08 |
Family
ID=71998690
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010414448.3A Active CN111544005B (en) | 2020-05-15 | 2020-05-15 | Parkinson's disease dyskinesia quantification and identification method based on support vector machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111544005B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112075940A (en) * | 2020-09-21 | 2020-12-15 | 哈尔滨工业大学 | Tremor detection system based on bidirectional long-time and short-time memory neural network |
| CN112674762A (en) * | 2020-12-28 | 2021-04-20 | 江苏省省级机关医院 | Parkinson tremble evaluation device based on wearable inertial sensor |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104398263A (en) * | 2014-12-25 | 2015-03-11 | 中国科学院合肥物质科学研究院 | Method for quantitatively evaluating symptoms of tremor of patient with Parkinson's disease according to approximate entropy and cross approximate entropy |
| CN105631195A (en) * | 2015-12-18 | 2016-06-01 | 合肥工业大学 | Wearable multi-information fusion gait analysis system and method thereof |
| CN106202952A (en) * | 2016-07-19 | 2016-12-07 | 南京邮电大学 | A kind of Parkinson disease diagnostic method based on machine learning |
| CN106821391A (en) * | 2017-03-23 | 2017-06-13 | 北京精密机电控制设备研究所 | Body gait acquisition analysis system and method based on inertial sensor information fusion |
| CN106971059A (en) * | 2017-03-01 | 2017-07-21 | 福州云开智能科技有限公司 | A kind of wearable device based on the adaptive health monitoring of neutral net |
| CN107103297A (en) * | 2017-04-20 | 2017-08-29 | 武汉理工大学 | Gait identification method and system based on mobile phone acceleration sensor |
| CN107958221A (en) * | 2017-12-08 | 2018-04-24 | 北京理工大学 | A kind of human motion Approach for Gait Classification based on convolutional neural networks |
| CN108309303A (en) * | 2017-12-26 | 2018-07-24 | 上海交通大学医学院附属第九人民医院 | A kind of wearable freezing of gait intellectual monitoring and walk-aid equipment |
| CN108932504A (en) * | 2018-07-24 | 2018-12-04 | 中国科学院深圳先进技术研究院 | Identity identifying method, device, electronic equipment and storage medium |
| CN108968918A (en) * | 2018-06-28 | 2018-12-11 | 北京航空航天大学 | The wearable auxiliary screening equipment of early stage Parkinson |
| CN109805935A (en) * | 2017-11-21 | 2019-05-28 | 北京周智物联科技有限公司 | A kind of intelligent waistband based on artificial intelligence hierarchical layered motion recognition method |
| CN110151189A (en) * | 2019-04-30 | 2019-08-23 | 杭州电子科技大学 | Non-linear gait dynamics method of discrimination for parkinsonian gait risk assessment |
| WO2019165207A1 (en) * | 2018-02-23 | 2019-08-29 | Loma Linda Universtiy | Systems and methods for detection and correction of abnormal movements |
| CN110327054A (en) * | 2019-07-17 | 2019-10-15 | 袁兴光 | A kind of gait analysis method and device based on acceleration and angular speed sensor |
| CN110647857A (en) * | 2019-09-29 | 2020-01-03 | 南京工程学院 | Gait recognition method and system |
-
2020
- 2020-05-15 CN CN202010414448.3A patent/CN111544005B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104398263A (en) * | 2014-12-25 | 2015-03-11 | 中国科学院合肥物质科学研究院 | Method for quantitatively evaluating symptoms of tremor of patient with Parkinson's disease according to approximate entropy and cross approximate entropy |
| CN105631195A (en) * | 2015-12-18 | 2016-06-01 | 合肥工业大学 | Wearable multi-information fusion gait analysis system and method thereof |
| CN106202952A (en) * | 2016-07-19 | 2016-12-07 | 南京邮电大学 | A kind of Parkinson disease diagnostic method based on machine learning |
| CN106971059A (en) * | 2017-03-01 | 2017-07-21 | 福州云开智能科技有限公司 | A kind of wearable device based on the adaptive health monitoring of neutral net |
| CN106821391A (en) * | 2017-03-23 | 2017-06-13 | 北京精密机电控制设备研究所 | Body gait acquisition analysis system and method based on inertial sensor information fusion |
| CN107103297A (en) * | 2017-04-20 | 2017-08-29 | 武汉理工大学 | Gait identification method and system based on mobile phone acceleration sensor |
| CN109805935A (en) * | 2017-11-21 | 2019-05-28 | 北京周智物联科技有限公司 | A kind of intelligent waistband based on artificial intelligence hierarchical layered motion recognition method |
| CN107958221A (en) * | 2017-12-08 | 2018-04-24 | 北京理工大学 | A kind of human motion Approach for Gait Classification based on convolutional neural networks |
| CN108309303A (en) * | 2017-12-26 | 2018-07-24 | 上海交通大学医学院附属第九人民医院 | A kind of wearable freezing of gait intellectual monitoring and walk-aid equipment |
| WO2019165207A1 (en) * | 2018-02-23 | 2019-08-29 | Loma Linda Universtiy | Systems and methods for detection and correction of abnormal movements |
| CN108968918A (en) * | 2018-06-28 | 2018-12-11 | 北京航空航天大学 | The wearable auxiliary screening equipment of early stage Parkinson |
| CN108932504A (en) * | 2018-07-24 | 2018-12-04 | 中国科学院深圳先进技术研究院 | Identity identifying method, device, electronic equipment and storage medium |
| CN110151189A (en) * | 2019-04-30 | 2019-08-23 | 杭州电子科技大学 | Non-linear gait dynamics method of discrimination for parkinsonian gait risk assessment |
| CN110327054A (en) * | 2019-07-17 | 2019-10-15 | 袁兴光 | A kind of gait analysis method and device based on acceleration and angular speed sensor |
| CN110647857A (en) * | 2019-09-29 | 2020-01-03 | 南京工程学院 | Gait recognition method and system |
Non-Patent Citations (2)
| Title |
|---|
| 乔子晏: "基于惯性传感器的PD患者运动功能评价的研究", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
| 刘遥: "基于惯性传感器的帕金森病上肢运动障碍量化评估方法研究", 《中国博士学位论文全文数据库 医药卫生科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112075940A (en) * | 2020-09-21 | 2020-12-15 | 哈尔滨工业大学 | Tremor detection system based on bidirectional long-time and short-time memory neural network |
| CN112674762A (en) * | 2020-12-28 | 2021-04-20 | 江苏省省级机关医院 | Parkinson tremble evaluation device based on wearable inertial sensor |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111544005B (en) | 2022-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yoneyama et al. | Accelerometry-based gait analysis and its application to Parkinson's disease assessment—part 1: detection of stride event | |
| Ying et al. | Automatic step detection in the accelerometer signal | |
| CN106228200B (en) | Action identification method independent of action information acquisition equipment | |
| Lee et al. | Parkinson’s disease classification using gait characteristics and wavelet-based feature extraction | |
| CN111544005B (en) | Parkinson's disease dyskinesia quantification and identification method based on support vector machine | |
| CN108451532B (en) | Method and system for acquiring respiratory information | |
| WO2013114596A1 (en) | Information processing device, method for generating representative waveform, and program for generating representative waveform | |
| CN108958482B (en) | Similarity action recognition device and method based on convolutional neural network | |
| CN111544006B (en) | Wearable equipment for quantifying and identifying dyskinesia of Parkinson's disease people | |
| Lahdenoja et al. | A smartphone-only solution for detecting indications of acute myocardial infarction | |
| CN111345814B (en) | Analysis method, device, equipment and storage medium for electrocardiosignal center beat | |
| Majumder et al. | A wireless IoT system towards gait detection in stroke patients | |
| CN110801227A (en) | Method and system for testing three-dimensional color block obstacle based on wearable equipment | |
| JP7489729B2 (en) | Method for preventing falls and device for carrying out such method | |
| Boutaayamou et al. | Algorithm for temporal gait analysis using wireless foot-mounted accelerometers | |
| CN116269413A (en) | Continuous electrocardiographic waveform reconstruction system and method using smart wristband motion sensor | |
| CN111345815B (en) | Method, device, equipment and storage medium for detecting QRS wave in electrocardiosignal | |
| Maziewski et al. | Accelerometer signal pre-processing influence on human activity recognition | |
| CN109692004B (en) | Method and device for detecting standing posture of human body | |
| Torres-Pardo et al. | Is Lyapunov exponent a reliable metric to detect dynamic stability in Parkinson’s disease? | |
| CN116942102B (en) | Sleep stage method and device based on pulse wave | |
| WO2019150905A1 (en) | Method, system and non-transitory computer readable medium | |
| CN111657918B (en) | Falling detection method and system combining electrocardio and inertial sensing data | |
| Vanitha et al. | Machine Learning Techniques for Automated Tremor Detection in the Presence of External Stressors | |
| CN210228151U (en) | Single-accelerometer human body gait detection device and wearable equipment |
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 |