KR102152957B1 - The discrimination of panic disorder from other anxiety disorders based on heart rate variability and the apparatus thereof - Google Patents

Abstract

The present invention relates to a method and a device for discriminating panic disorder from other anxiety disorders by machine learning and based on heart rate variability (HRV). The method includes the following steps of: (S100) receiving an HRV signal for a preset time; (S200) extracting an HRV feature with respect to the received HRV signal; (S300) performing learning by inputting the extracted HRV feature to a machine learning algorithm; (S400) determining training and test data sets used in the learning step; (S500) dividing the training data set into lower training and validation sets; (S600) performing cross validation by using the lower training and validation sets; and (S700) comparing the cross validation execution result value with a preset value.

Description

The discrimination of panic disorder from other anxiety disorders based on heart rate variability and the apparatus thereof

The present invention relates to a method and apparatus for discriminating panic disorder, and more particularly, to a method and apparatus for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV).

In psychiatry, panic disorder and other anxiety disorders differ greatly in their major clinical characteristics, therapeutic approaches, and prognosis. However, in the clinical field, it is difficult to quickly and accurately diagnose depressive disorder and anxiety disorder due to the limited number of treatments and duration of treatment. In particular, in situations where legal issues such as industrial accidents, traffic accidents, military service, etc. are connected, a test that can judge panic disorder and other anxiety disorders is necessary in addition to the subjective symptoms of patients.

Other methods commonly used in clinical settings include Magnetic Resonance Imaging (MRI) and Electroencephalogram (EEG). However, in the case of anxiety disorder, damage to the brain parenchyma is not often accompanied, so it is impossible to visually check the brain magnetic resonance imaging test results and use them for discrimination.

On the other hand, by analyzing the results obtained from the brain magnetic resonance imaging test, it is possible to confirm a minute change through the measured value of the voxel unit or cortical thickness. However, this approach is also difficult to apply to distinguishing panic disorder from anxiety disorder. In addition, since the cost of a brain magnetic resonance imaging test is quite high, there is a limit that cannot be used frequently in clinical settings.

On the other hand, electroencephalography is used for diagnosis of anxiety disorders because it is cheaper than brain magnetic resonance imaging and has the advantage of not having any discomfort during the test.

However, the EEG test also has a disadvantage that the measurement time is quite long, there is a difference between the arousal EEG and the sleep EEG, and the result is different depending on the number of electrode channels of the EEG measurement device. There are limits to application.

Meanwhile, the heart rate variability (HRV) is a change in heart rate over time, and the state of the autonomic nervous system and the degree of response to stress can be predicted using the heart rate variability. When an abnormality occurs in the autonomic nervous system, the width of the heart rate variability decreases, and the body cannot adequately respond to stress and becomes vulnerable to various diseases. Since the method of measuring heart rate variability (HRV) is non-invasive, it is easy to implement and has a characteristic that the burden on the measurement subject is low.

(US) Patent Publication No. 2017-0238858

The present invention has been devised to solve the above problem, and an object of the present invention is to provide a method and apparatus for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV).

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems that are not mentioned will be clearly understood by those skilled in the art from the following description.

A method for discriminating panic disorder from other anxiety disorders according to an embodiment of the present invention includes the steps of: receiving, by the panic disorder discrimination apparatus 1000, a heart rate variability (HRV) signal for a preset time (S100); Extracting a heart rate variability (HRV) feature from the received heart rate variability (HRV) signal (S200); Inputting and learning the extracted heart rate variability (HRV) feature into a machine learning algorithm (S300); Determining a training data set and a test data set used in the learning step (S400); Dividing the training data set into a lower training set and a verification set (S500); It may include performing cross-validation using the sub-training set and the verification set (S600), and comparing the cross-validation result value with a preset value (S700).

An apparatus for discriminating panic disorder from other anxiety disorders according to an embodiment of the present invention includes: a heart rate variability signal receiving unit 100 for receiving a heart rate variability (HRV) signal for a preset time; A feature extraction unit 200 for extracting a heart rate variability (HRV) feature with respect to the received heart rate variability (HRV) signal; A machine learning unit 300 for learning by inputting the extracted heart rate variability (HRV) feature into a machine learning algorithm; A data set determination unit 400 that determines a training data set and a test data set used by the machine learning unit, and divides the training data set into a lower training set and a verification set; And a cross-validation unit 500 that performs cross-validation using the lower training set and the verification set. And a comparison unit 600 comparing the cross-verification result value with a preset value.

For the treatment of anxiety disorders, patients must take appropriate antidepressants or anti-anxiety drugs such as selective serotonin reuptake inhibitors (SSRIs) for a considerable period of time, and psychological and psychosocial treatments are also required. Anxiety disorders often take a long time for a patient to recover completely, or they cannot recover completely and go through a process of becoming chronic due to persistent symptoms.

The method and apparatus for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV) according to an embodiment of the present invention have the advantage of being able to accurately and quickly perform differential diagnosis of anxiety disorders and panic disorders.

Therefore, there is an advantage of being able to quickly treat an appropriate treatment for a patient by using the present invention.

The effects of the present invention are not limited to the above-mentioned effects, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a block diagram of an apparatus for discriminating a panic disorder from other anxiety disorders according to an embodiment of the present invention.
2 is a flow chart of a method for discriminating a panic disorder from other anxiety disorders according to an embodiment of the present invention.

In the present invention, various modifications may be made and various embodiments may be provided, and specific embodiments will be described in detail with reference to the drawings. However, this is not intended to limit the present invention to a specific embodiment, it is to be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present invention. In describing each drawing, similar reference numerals have been used for similar elements.

Terms such as first, second, A, and B may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of discriminating one component from another component. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component. The term and/or includes a combination of a plurality of related items or any of a plurality of related items.

When a component is referred to as being "connected" or "connected" to another component, it should be understood that it may be directly connected to or connected to the other component, but other components may exist in the middle. something to do. On the other hand, when a component is referred to as being "directly connected" or "directly connected" to another component, it should be understood that there is no other component in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present application, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof does not preclude in advance.

In addition, it is to be understood that the terms "first" and "second" are used herein for discrimination purposes only, and are not meant to represent or predict sequences or priorities in any way.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in this application. Does not.

Throughout the specification and claims, when a certain part includes a certain component, it means that other components may be further included rather than excluding other components unless otherwise stated.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of an apparatus for discriminating a panic disorder from other anxiety disorders according to an embodiment of the present invention.

Referring to FIG. 1, an apparatus 1000 for discriminating a panic disorder from other anxiety disorders according to an embodiment of the present invention includes a heart rate variability signal receiving unit 100; A feature extraction unit 200; Machine learning unit 300; A data set determination unit 400, a cross-verification unit 500, and a comparison unit 600 may be included.

The heart rate variability signal receiving unit 100 according to an embodiment of the present invention may receive a heart rate variability (HRV) signal for a preset time, and may be converted into a heart rate variability by receiving an electrocardiogram signal (ECG) through an electrode. have.

More specifically, heart rate variability (HRV) can be detected from pulse waveforms appearing on an electrocardiogram (ECG) signal, blood pressure, and blood volume meter. Among them, the method of checking heart rate variability (HRV) from an electrocardiogram signal is the most. It is known to be accurate. At this time, the electrocardiogram (ECG) signal may be converted into a heart rate variability (HRV) by a QRS detection algorithm. At this time, the QRS wave (QRS complex) is a waveform that is displayed by the depolarization process of the ventricular muscle, P, It refers to a waveform having five curves named Q, R, S, and T waves.

The feature extraction unit 200 according to an embodiment of the present invention may extract a heart rate variability (HRV) feature from the heart rate variability (HRV) signal received by the heart rate variability signal receiving unit 100.

Heart Rate Variability (HRV) features are divided into a high frequency (HF) domain between 0.15 Hz and 0.4 Hz, a low frequency (LF) between 0.04 Hz and 0.15 Hz, and a very low frequency (VLF) band between 0.0033 Hz and 0.04 Hz, depending on the frequency. SDNN (Standard-to-Normal interval) standard deviation, Root Mean Square of the Successive Differences (RMSSD), Physical Stress Index (PSI), Pressure Index, Total Power Index (TP), etc. It may include. The characteristics of heart rate variability will be described later.

The machine learning unit 300 according to an embodiment of the present invention may learn by inputting the extracted heart rate variability (HRV) feature into a machine learning algorithm, and the machine learning algorithm is a random forest, a gradient boosting machine ( Gradient Boosting Machine), Support Vector Machine, Artificial Neural Network, and normalized logistic regression analysis may be included.

The data set determination unit 400 according to an embodiment of the present invention determines a training data set and a test data set used by the machine learning unit 300, and the training data set is a lower level. It can be divided into a training set and a validation set.

The cross-validation unit 500 according to an embodiment of the present invention performs cross-validation using the lower training set and the verification set, and the comparison unit 600 is a cross-validation result value, that is, a result of a machine learning prediction model. When the value is greater than a predetermined value by comparing the value with a preset value, it may be determined as a panic disorder. However, the criterion for determining panic disorder is not limited to a case that is greater than or equal to a predetermined value compared to a preset value, and may be less than a predetermined value, and may be variously set according to embodiments.

2 is a flow chart of a method for discriminating a panic disorder from other anxiety disorders according to an embodiment of the present invention.

The method for discriminating panic disorder from other anxiety disorders according to an embodiment of the present invention is a method performed by the panic disorder discrimination apparatus 1000, and receiving a heart rate variability (HRV) signal for a preset time (S100) ); Extracting a heart rate variability (HRV) feature from the received heart rate variability (HRV) signal (S200); Inputting and learning the extracted heart rate variability (HRV) feature into a machine learning algorithm (S300); Determining a training data set and a test data set used in the learning step (S400); Dividing the training data set into a lower training set and a verification set (S500); It may include performing cross-validation using the sub-training set and the verification set (S600), and comparing the cross-validation result value with a preset value (S700).

Hereinafter, an embodiment of performing a method for discriminating a panic disorder from other anxiety disorders according to an embodiment of the present invention will be described.

Participant

HRV data was obtained from medical charts recorded from January 2011 to December 2017. The data inclusion criteria are as shown in Table 1 below, (1) between the ages of 20-65, (2) Panic Disorder (10th edition of the International Statistical Classification Illness and Related Health Problems (ICD-10) Code for F41.0) or Patients diagnosed with other anxiety disorders (F41.1 to F41.9) and (3) during the first month of hospital visit.

ICD-10-CM Codes Number F41.0 Panic disorder 60 F41.1 Generalized anxiety disorder 8 F41.3 Other mixed anxiety disorders 23 F41.8 Other specified anxiety disorders 18 F41.9 Anxiety disorder, unspecified 12

The data exclusion criteria are (1) clinically significant cardiovascular diseases such as arrhythmia and cardiomyopathy and (2) co-psychiatric and material disorders other than anxiety disorders.

HRV

Heart rate variability measurement was performed by a doctor using BFM-5000PLUS (Medicore; Seoul, Korea). Prior to testing, subjects were instructed to describe and follow the following precautions. On the day before the measurement, excessive alcohol consumption or vigorous exercise was not allowed, caffeine drinks, cigarettes or drugs were not consumed for 3 hours before the test, and food was not consumed from 2 hours before. On the day before the test, he was advised to get enough sleep and refrain from behaviors that could cause fatigue. Since heart rate changes represent daily changes, measurements were taken from 8 a.m. to 12 a.m., and the laboratory maintained a bright and soft light, shielded out external noise, and maintained an appropriate room temperature (20-25 °C).

The subject could not wear accessories such as precious metals that could affect the ECG waveform, and the subject attached three electrodes to each of the left and right wrists and left ankles while sitting in a chair. After 5 minutes or more of preparation time was applied to sufficiently adapt to the test environment, the heart rate change was measured for 5 minutes. During the examination, the subjects remained motionless and speechless, with their eyes open, sitting in a comfortable mind and posture, and breathing comfortably.

The whole process of machine learning

The ratio of training and hold-out test set was determined to be 0.7 and 0.3, respectively. The training set was again divided into sub-training set and validation set into 0.7 (0.49 of the total data set) and 0.3 (0.21 of the total data set). Hyperparameters (hyperparameters) were adjusted through 10-fold cross-validation with sub-training set and validation set. The performance of the final predictive model from cross-validation was tested with a hold-out test set that was never used. The whole process of machine learning was done using Python's Scikit Learning Library ('sklearn') in 3.5.

Predictive model and algorithm

In an embodiment of the present invention, logistic regression (LR), artificial neural network (ANN), gradient boosting machine (GBM), random forest (RF), and support vector machine ( A total of five algorithms, such as support vector machine (SVM) were used. The main outcome measures for the performance of the predictive model are the overall accuracy and F ₁ score.

Feature Selection

The main frequency band of the HRV signal is between 0 (or 0.2) and 0.4 Hz and can be clinically divided into two frequency domains: a high frequency (HF) domain and a low frequency (LF) domain. The standard for division is the high-frequency range of 0.15-0.4Hz and the low-frequency range of 0.04-0.15Hz. And very low frequency (VLF) is a frequency band of 0.0033-0.04 Hz, and very low frequency (VLF) is known to provide additional information for most sympathetic nerves. The main frequency bands of very low frequency (VLF), low frequency (LF) and high frequency (HF) and their power distribution are not static and may vary depending on the respiratory and cardiac control of the autonomic nervous system.

The change in the high frequency (HF) region is related to the parasympathetic activity of the vagus nerve according to respiration, and the variability size of the low frequency (LF) region can be affected by the vagus nerve activity and the sympathetic nerve activity. The ratio of low frequency (LF)/high frequency (HF) is an indicator that reflects the balance between the activity of the sympathetic nerve and that of the parasympathetic nerve.

The standard deviation of SDNN (Standard-to-Normal interval) refers to the stress resistance. This feature is the standard deviation of the total normal-normal interval, indicating how much the heart rate changes during the recording time.

RMSSD (Root Mean Square of the Successive Differences) is the square root of the mean square of the difference between adjacent RR intervals in heart rate variability (HRV) and is used to evaluate the high frequency region. This feature represents the parasympathetic regulation of the heart.

Physical Stress Index (PSI) refers to the stress index of the automatic analysis result paper and the degree of pressure applied to the regulatory system using parameters such as HR + SDN + RMSS in the time domain. The approximate entropy (ApEn) refers to the complexity of the signal, which represents a statistical quantification of how complex the signal is in the time domain. SRD (Successful RR Interval Difference) indicates the reliability of the test. The degree of change is measured based on a specific segment at 5-minute intervals, and a value of 1 indicates whether this data remains constant.

Total Power (TP) refers to the sum of autonomic activity and all power sources, including very low frequency (VLF), low frequency (LF) and high frequency (HF). This reflects the overall activity of the autonomic nervous system, which indicates its ability to control the autonomic nervous system.

Table 2 below shows information on the demographic and heart rate variability (HRV) of patients with panic disorder and other anxiety disorders. There were no significant differences in age and sex between panic disorder and other anxiety disorder groups.

Panic Other anxiety Statistics P value Age 42.7 (11.8) 49.3 (10.4) 3.265 0.001 Sex, female 29 (48.3) 27 (44.3) 0.202 0.653 Mean HR 78.2 (11.8) 74.5 (13.7) -1.453 0.149 SDNN 35.0 (16.5) 37.9 (25.1) 0.739 0.462 RMSSD 25.5 (14.9) 27.5 (19.5) 0.627 0.532 PSI 74.1 (98.3) 80.0 (95.6) 0.336 0.738 ApEn 1.1 (0.1) 1.1 (0.2) -2.652 0.009 SRD 0.9 (0.1) 0.9 (0.1) -0.869 0.387 TSRD 107.8 (40.0) 112.1 (45.9) 0.543 0.588 TP 1121.0 (1257.8) 1199.0 (1812.5) 0.275 0.784 VLF 491.9 (645.2) 725.1 (1280.6) 1.262 0.210 LF 410.4 (522.8) 276.7 (421.0) -1.551 0.124 HF 218.8 (258.5) 197.3 (259.0) -0.458 0.648 LF/HF 2.40 (2.23) 2.31 (3.17) -0.177 0.860 BAI 30.2 (13.7) 21.5 (11.3) -3.815 0.000

All variables are expressed as mean (SD) for continuous type and N (%) for categorical variable.

Table 3 below shows the performance of a machine learning-based classification model for panic disorder and other anxiety disorders.

RF GBM SVM ANN LR Accuracy 0.649 0.676 0.730 0.730 0.784 Sensitivity 0.667 0.722 0.722 0.833 0.833 Specificity 0.632 0.632 0.737 0.632 0.737 Precision 0.632 0.650 0.722 0.682 0.750 NPP 0.667 0.706 0.737 0.800 0.824 FPR 0.368 0.368 0.263 0.368 0.263 FDR 0.368 0.350 0.278 0.318 0.250 FNR 0.333 0.278 0.278 0.167 0.167 F1-score 0.649 0.684 0.722 0.750 0.790 MCC 0.298 0.355 0.460 0.473 0.572

In Table 3, NPP (Negative Predictive Value) means negative predictive value, FPR (False Positive Rate) represents false positive rate, FDR (False Discovery Rate) represents false discovery rate, and FNR (False Negative Rate) represents false negative rate. , MCC (Matthews Correlation Coefficient) means the Matthews correlation coefficient.

Referring to Table 3, the normalized logistic regression (LR) has the best classification performance. Normalized logistic regression (LR) was shown not only for overall accuracy, but also for MCC and F1 scores.

The hyperparameters (hyperparameters) of the five machine learning algorithms are shown in Table 4. More specifically, for the main hyperparameters of normalized logistic regression, the L1 penalty was applied as a'liblinear' solver. Tolerability was set to 0.0001, C was set to 0.3, and the maximum number of repetitions was set to 100.

Random Forest n_estimators = 500, criterion = "entropy", max_depth = None, min_samples_split = 2, min_samples_leaf = 1, min_weight_fraction_leaf = 0, max_features = "auto", max_leaf_nodes = None, min_impurity_decrease = 0, min_impurity_decrease = 0, min_impurity_decrease = 0, min_impurity_decrease = 0, min_impurity, bootstrap = 0, min_impurity, boots, split = 0.000000 False, verbose = 0 Gradient boosting machine n_estimators = 500, learning_rate = 0.01, loss = exponential, subsample = 0.8, min_samples_split = 10, min_samples_leaf = 5, max_depth = 3, min_impurity_decrease = 2.0, min_impurity_split = 0.0000007, max_features = 2, max_leaf_nodes = None, validation_fraction Support vector machine C = 0.21, gamma = 0.0645, kernel = "poly", degree = 2, coef0 = 0.6, shrinking = True, probability = False, lass_weight = 1:0.844, decision_function_shape ='ovr', tol = 0.003, verbose = False, max_iter = -1 Artificial neural network max_iter = 1000, hidden_layer_sizes = 100, activation = "tanh", solver = "sgd", alpha = 0.0002, batch_size = "auto", learning_rate ='constant', learning_rate_init = 0.001, shuffle = True, momentum = 0.1, nesterovs_momentum = True, tol = 0.0001, validation_fraction = 0.1 Regularized logistic regression solver = "liblinear", penalty = "l1", dual = false, tol = 0.0001, C = 0.3, intercept_scaling = 1, fit_intercept = True, max_iter = 100, verbose = 0

Taken together, the present invention proposes a method of distinguishing panic disorder from other anxiety disorders. As a result of the analysis, LR showed the best performance with accuracy of 0.785 and F1-score of 0.790, respectively. Therefore, the optimal embodiment of the present invention is the case of using a logistic regression (LR) algorithm.

Although several studies have shown that anxiety disorders have an abnormal pattern of HRV, these studies have only been presumed to confirm the physiological status of anxiety disorders.

The results of the embodiments of the present invention can be applied to a variety of situations and institutions, such as community mental health centers, primary health institutions, and work or school based mental health care. The present invention can be said to have strengths in the fields of clinical psychiatry and public mental health systems.

In summary, the present invention is expected to be useful to health workers in various institutions in classifying panic disorder and other anxiety disorders.

The above description is merely illustrative of the technical idea of the present invention, and those of ordinary skill in the technical field to which the present invention pertains will be able to make various modifications and variations without departing from the essential characteristics of the present invention. Accordingly, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention, but to explain the technical idea, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be interpreted as being included in the scope of the present invention.

100: heart rate variability signal receiver
200: feature extraction unit
300: Machine Learning Department
400: data set determination unit
500: cross verification unit
600: comparison unit
1000: Panic disorder detection device

Claims (10)

A device for discriminating panic disorder from other anxiety disorders,
Receiving a heart rate variability (HRV) signal for a preset time;
Extracting a heart rate variability (HRV) feature from the received heart rate variability (HRV) signal;
Learning by inputting the extracted heart rate variability (HRV) feature into a machine learning algorithm;
Determining a training data set and a test data set used in the learning step;
Dividing the training data set into a sub training set and a verification set; And
Performing cross-validation using the sub-training set and the verification set,
Further comprising the step of comparing the cross-validation result value with a preset value,
The cross-validation is characterized in that the hyperparameter is adjusted and the cross-validation is performed,
The heart rate variability (HRV) feature is a method for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV), characterized in that it includes a physical stress index (PSI) and a total power index (TP).
delete The method of claim 1,
The machine learning algorithm is at least one of a random forest, a gradient boosting machine, a support vector machine, an artificial neural network, and a normalized logistic regression analysis. How to differentiate panic disorder from other anxiety disorders based on heart rate variability (HRV). The method of claim 1,
The heart rate variability (HRV) signal is divided into a high frequency (HF) domain between 0.15 Hz and 0.4 Hz, a low frequency (LF) between 0.04 Hz and 0.15 Hz, and a very low frequency (VLF) band between 0.0033 Hz and 0.04 Hz, depending on the frequency. A method for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV), characterized in that extracting heart rate variability (HRV) features. The method of claim 1,
The heart rate variability (HRV) feature further comprises at least one of SDNN (Standard-to-Normal Interval) standard deviation and Root Mean Square of the Successive Differences (RMSSD). How to differentiate panic disorder from other anxiety disorders based on variability (HRV). In the device for discriminating panic disorder from other anxiety disorders,
A heart rate variability signal receiving unit 100 for receiving a heart rate variability (HRV) signal for a preset time;
A feature extraction unit 200 for extracting a heart rate variability (HRV) feature from the received heart rate variability (HRV) signal;
A machine learning unit 300 for learning by inputting the extracted heart rate variability (HRV) feature into a machine learning algorithm;
A data set determination unit 400 that determines a training data set and a test data set used by the machine learning unit, and divides the training data set into a lower training set and a verification set; And
Including a cross-validation unit 500 for performing cross-validation using the lower training set and the verification set,
Further comprising a comparison unit 600 for comparing the cross-validation result value with a preset value,
The cross-validation is characterized in that the hyperparameter is adjusted and the cross-validation is performed,
The heart rate variability (HRV) feature is a device for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV), characterized in that it includes a physical stress index (PSI) and a total power index (TP).
delete The method of claim 6,
The machine learning algorithm is at least one of a random forest, a gradient boosting machine, a support vector machine, an artificial neural network, and a normalized logistic regression analysis. Panic disorder discrimination device from other anxiety disorders based on heart rate variability (HRV). The method of claim 6,
The heart rate variability (HRV) signal is divided into a high frequency (HF) domain between 0.15 Hz and 0.4 Hz, a low frequency (LF) between 0.04 Hz and 0.15 Hz, and a very low frequency (VLF) band between 0.0033 Hz and 0.04 Hz, depending on the frequency. A device for discriminating panic disorder from other anxiety disorders based on heart rate variability (HRV), characterized in that the heart rate variability (HRV) feature is extracted. The method of claim 6,
The heart rate variability (HRV) feature further comprises at least one of SDNN (Standard-to-Normal Interval) standard deviation and Root Mean Square of the Successive Differences (RMSSD). A device for discriminating panic disorder from other anxiety disorders based on variability (HRV).

Images