TWI442348B - Personalized gait analysis method of physiological sensor - Google Patents

Description

Personalized gait analysis method for physiological sensors

The invention relates to a personalized physiological sensor with biofeedback, in particular to an automatic analysis of a user's exercise habits by personalization, analysis of gait, and adaptation of different exercise states according to different users. Music, and when the exercise exceeds the intensity of the load of the heart, through the voice prompts of the headphones, immediately give a voice suggestion to suspend the intense exercise, to avoid the potentially dangerous physiological sensor of the excessively intense physical behavior of the personalized gait analysis method.

Hypertension, diabetes, hyperlipidemia and overweight are the main risk factors for cardiovascular disease, and the mortality rate of cardiovascular disease has been one of the top ten causes of death among Chinese people. It is one of the urgent problems to be solved. In particular, lethal heart disease often occurs without warning, and once it happens, it usually causes regrets for the world. In order to encourage the movement of the sport, in recent years, under the government's vigorous promotion of the "sports population multiplication plan" and the increase in people's health awareness, China's sports population has increased year by year. According to the results of a survey of national sports behaviors conducted by the National Health Service of the Health Department of the Executive Yuan, the most common exercise for Chinese people is walking (including walking), followed by jogging.

But the general public is not clear, although exercise helps to improve heart and lung function, but too intense exercise behavior is potentially dangerous. Sudden death in young people is often a defect in cardiovascular anatomy. It is induced during intense exercise. Through the heartbeat and recovery of the movement before and after exercise, the heart function can be easily evaluated and the possible dangers can be discovered in advance. In order to avoid the above regrets, the development of personalized physiological monitoring smart sensors is very important.

The measurement of the traditional exercise heart rate is affected by the body movements during exercise, so a special heart rate measuring instrument is needed. When performing exercise heart rate measurement without the assistance of special instruments, only It can be represented by the heart rate at the end of the exercise. According to the Ministry of Education's information published on fitness, the heart rate measurement of heartbeat is closely related to exercise intensity. Through the reaction of heartbeat during exercise, we can understand the physical load during exercise. The intensity during exercise is preferably 60 to 80% of the maximum heart rate.

Therefore, medical and health care has brought new impacts in the field of electronic information, and information technology is gradually changing the medical and health industry, further changing the traditional medical model, providing health care applications in the future, and joining the audio-visual application. The elements of entertainment, and the application of bio-feedback systems, provide users with a more user-friendly mode of use, thereby changing lifestyle patterns and reducing the incidence of cardiovascular disease.

The main object of the present invention is to provide a personalized gait analysis method for a physiological sensor, which can automatically analyze the user's exercise habits, analyze the gait by personalization, and can load the subject beyond the load of the heart. When the heart rate is reached, the reminder is given immediately by the voice, so that the person in motion can grasp the heart rhythm condition at any time and ensure the safety of the exercise.

Therefore, in order to achieve the above objective, the physiological sensor of the personalized biofeedback disclosed in the present invention comprises an electrocardiogram signal acquisition module, a three-axis accelerator sensor module, an analog-to-digital signal conversion module, a microprocessor, And a voice playback module. The electrocardiogram signal capture module and the three-axis accelerator sensor module respectively detect the electrocardiogram signal and the motion signal of the subject, and the microprocessor receives the action signal after being converted into the digitized signal by the analog-to-digital digital signal conversion module. The "quartile" method is used to judge the motion state of the subject, and the critical physiological state signal of the motion state and the reference is calculated according to the quartile algorithm, and the critical physiological state signal and the electrocardiogram signal are compared. If the ECG signal exceeds the critical physiological state signal range, the notification sound is played to remind the subject to avoid the potential danger of excessively intense exercise behavior.

On the other hand, the physiological sensor can further include a storage module, which can store the digitized ECG signal and the three-axis motion signal, and transmit it to the computer system for analysis through the wireless transmission module, thereby establishing personalization. The information is then fed back to the sensor to match the various motion states to enable the voice module to play the corresponding music, and the microprocessor can make a reference for the judgment.

At the same time, the satellite positioning module can be used to detect the motion path of the subject, and the average heart rate, the calorie consumption accumulation, and the exercise distance accumulation of the subject are calculated.

In order to further understand the purpose, features and functions of the present invention, the drawings are described in detail as follows:

As shown in the first figure, a schematic diagram of a physiological sensor with biofeedback provided by an embodiment of the present invention.

The physiological sensor of the personalized biofeedback disclosed in the present invention comprises a microprocessor 101, a three-axis accelerator sensor module 102, an electrocardiogram signal acquisition module 103, a satellite positioning module 104, and an analog-to-digital signal. The conversion module 105, the voice playback module 106, the battery module 107, the storage module 108, the display module 109, the wireless transmission module 110, and the input module 111.

The electrocardiogram signal capture module 103 is used to extract the electrocardiogram signal of the subject. The muscle of the heart is the only muscle in the human muscle that has spontaneous beat and rhythmic contraction, and the electrocardiogram (ECG) records the voltage change of the heart tissue. The electrical signal, the ECG's frequency is below 150Hz (also there is a 250Hz), the signal size is about 1mV, to process the ECG signal, the ECG must be enlarged 1200 Double the size to facilitate handling.

The heart can be divided into the left atrium, the left ventricle, the right atrium, and the right ventricle. The contraction of the heart is controlled by a small electrical impulse of about 60 times per minute produced by the right atrial sinus node (SA node). Generally, the heart cells in a static state are charged (negatively charged), or "polarized". Once they are electrically stimulated, they are "deporlarized", positively charged and produce a contractile response. Although the heart can operate independently from the sinoatrial node, it is possible to borrow the sympathetic nerve (stimulation contraction) and the parasympathetic nerve (sedation) to speed up or lower the blood by transmitting signals from the brain and various parts of the body to assist the heart to adjust the heart rate. The cycle speed is taken in response to various external situations.

The triaxial accelerator sensor module 102 is mainly used to measure the limb movement state of the subject, and obtain the motion state signal, and use the "quartile" method to discriminate the various motion states of the subject, such as being stationary. Physical activity such as walking and running. Since each person runs and walks in different ways, the three-axis accelerator sensor module 102 can be used to detect signals of various sports postures such as personalized running, and then the parameters after learning can be used as subsequent analysis. Basis.

The satellite positioning module 110 is configured to detect the position signal of the subject, which is a Global Position System (GPS), which is composed of 24 artificial satellites using a satellite navigation system of a global scale. Using the C/A code standard measurement method open to the public, the accuracy of tens of meters can be obtained. After the satellite receiver is positioned, the effective positioning data is transmitted through the output pipeline, including the following longitude, latitude, positioning completion code, the number of effective satellites used, the satellite number used, and the elevation angle, direction angle, received signal strength, Satellite azimuth, high Degree, relative displacement displacement velocity, relative displacement displacement direction angle, date, UTC time, DOP error reference value, satellite status and reception status. Therefore, it is possible to provide information about the motion of the subject, the average heart rate, the calorie consumption accumulation, and the exercise distance accumulation.

The analog-to-digital digital signal conversion module 105 is configured to convert the motion signal sensed by the three-axis accelerator sensor module 102 and the electrocardiogram signal detected by the electrocardiogram signal acquisition module 103 into a digitized signal for subsequent processing. The most common conversion modes are single channel, sequential channel, repeated single channel, repeated sequential channel, and so on.

The wireless module 110 can be used for communication between the physiological sensor and the computer system. Generally, the standard is a short-range radio frequency wireless connection technology using Bluetooth, which is mainly used to provide short distance, low cost and low. Power-consuming wireless network communication transmission, because it can be used in the 3C field of information, communication and consumer electronics, to provide mobile phone voice, signal data, video and other transmission functions, has been widely used in On a variety of portable devices.

The storage module 108 is configured to store an action signal sensed by the three-axis accelerator sensor module 102, an electrocardiogram signal detected by the electrocardiogram signal capture module 103, and a position signal detected by the satellite positioning module 104, thereby providing Calculate the motion information of the subject's motion path, average heart rate, calorie expenditure accumulation, and exercise distance accumulation. Of course, the voice data to be played by the voice playing module 106 can also be stored. In general, the most common storage device for Micro Secure Digital (also known as Trans Flash) is only 15×11×1mm (0.59×0.43×0.04 inches), which is a flash memory and Control the storage device composed of the wafer.

The battery module 107 is used to provide a module and the power required by the microprocessor 101 to perform operations. The force can be charged by a rechargeable lithium battery, and can be charged by, for example, a USB or a voltage device.

The display module 109 is used to display various information of the physiological sensor, and at the same time, as an interface for communicating with the subject, the most common one is a liquid crystal display (LCD).

The input module 111 is for the user to input an input message, such as time, height, weight and control menu, and the input module 111 is connected to the microprocessor 101. The input module 111 is generally the most common button or shake. Rod.

The voice playing module 106 has a music decoding IC, which can support music formats such as MPEG1.0 & 2.0 audio layer III (CBR, VBR, ABR); layers I & II optional; WAV (PCM + IMA ADPCM), built-in high The quality stereo DAC can directly stream the decoded MP3 music data (Bitstream) and output the sound through the stereo headphones. In addition, the voice playing module further includes a recording unit, and the subject can record personalized voice and music.

Therefore, during the exercise, the subject can be played by the voice play module 106 by reading the music file from the storage module 108, and at the same time, the posture of the exercise ( intense or tempered) can be changed according to different users. Appropriate music, so that the style of the music during the exercise can match the sporting style of the time, in order to increase the willingness to use and exercise time.

In use, the subject can store the relevant information obtained in the exercise state in the storage module 108 by using the advanced learning mode, and transmit the data to the computer system for analysis by using the wireless transmission module 110. It can correctly judge the movement state corresponding to the user, and set various personalized information, including instant heart rate calculation, human body activity posture judgment (walking or running), and timely feedback of heart rate and warning, which can be automatically Detect the current active posture and heart rate to make the choice of automatic music style. At the same time, because of the satellite positioning module 104, the motion is recorded in detail. The coordinate position and path of the candidate are to be returned to the personal computer by the end of the exercise, and the physiological data can be uploaded by the wireless transmission module 110 for the computer system to automatically generate various information reports of the exercise. , which can include calorie consumption, heart rate, motion path, time spent, average rate, etc., combined with Google Earth to indicate the various punctuation points and the current physiological conditions (such as average heart rate, calorie consumption accumulation, exercise distance accumulation), let use You can easily review each exercise process to set the next exercise goal.

On the other hand, during the motion, the three-axis accelerator sensor module 102 senses the motion signal of the subject at any time, and according to the previously learned motion state corresponding signal, can correctly determine the motion state of the subject, and The algorithm calculates the critical physiological state signal based on the motion state, and the ECG signal acquisition module 103 monitors the physiological state of the subject at any time. Once the ECG signal exceeds the critical physiological state signal, the original is cut off. The music played by the voice play module 106 is instead played to announce the voice to remind the subject to avoid the potential danger of excessively intense exercise behavior.

Therefore, the physiological sensor with biofeedback provided by the present invention immediately gives a voice suggestion for suspending intense exercise through the voice prompt of the earphone when the exercise exceeds the intensity of the load that can be carried by the heart. This study also combines the GPS global satellite positioning system to record in detail the coordinates and path of the motion. When the user's exercise ends and returns to the PC, various physiological data can be uploaded by the Bluetooth connection to make various applications. It is hoped that the user who provides the exercise through this device can have safe, pleasant and good sports experience. Let preventive medicine combine entertainment and enhance the added value of intelligent physiological monitoring instruments and medical equipment.

Please refer to the second figure, which is an electrocardiogram (ECG) of the present invention. A schematic diagram of the signal. The so-called cycle of heart beat is the electrical pulse from the sinus node transmitted to the left and right atrium in a progressive wave, causing the contraction of the left and right atrium (point P in Figure 2). After the AV node, it is about 1/10 second. This 1/10 second is to let the blood flow to the ventricle. Then the electric pulse transmits the electric pulse by transmitting the fiber (Q point in Fig. 2). The left and right ventricles contract to the left and right ventricles (point R in Fig. 2). After a series of electrical activities, the heart is temporarily still, and the ventricles wait for repolarization to restore the negatively charged state (point T in Fig. 2) to complete a heartbeat. . The ventricular depolarization and repolarization phenomena are the Q, R, S, and T points in Fig. 2, while the atrium has only the depolarized P point, and there is no repolarized waveform. This is because the atrial repolarization The waveform of the phenomenon is small and mostly masked in the composite wave at the Q, R, and S points, so it is not easy to detect. Therefore, the P, Q, R, S, and T points can be easily detected by checking the signal signal by the electrocardiogram.

After sampling the unfiltered ECG signal, the bandpass filter consisting of high-pass and low-pass filters is used to enhance the QRS complex. After the ECG signal is filtered, the characteristics of the QRS complex can be more prominent by the differentiator. The R wave can be clearly distinguished, while the relative P and T waves are again reduced. Then, the obtained electrocardiogram is squared in absolute value, and the signal is characterized by a window averaging method, and the electrocardiogram signal can be converted into a signal similar to a square wave.

Please refer to the third figure, which is a flow chart of the steps of the personalized gait analysis method of the physiological sensor of the present invention. The test subject can input a predetermined learning mode through an input module (for example, a button group), and the three-axis accelerator sensor module detects the motion signals of the personalized running and walking various motion postures, and the microprocessor obtains the motion. The signal uses the quartile algorithm to determine the various motion states of the subject. The flow chart of the steps of the personalized gait analysis method includes: Step S30: input a preset learning mode. If the preset time is 10 seconds, the subject can start the motion action; and step S31: read the learning time required by the subject.

Step S32: determining whether the learning time of the subject reaches the preset learning time. If not, executing step S33, and if yes, executing step S34.

Step S33: Store the waveform values of a three-axis accelerator module in an array (Array List), and perform step S31 again.

Step S34: sorting the waveform values stored in the array from low to high increments, wherein the three-axis accelerator module generates waveforms of three-axis accelerations in the X-axis, the Y-axis, and the Z-axis according to the motion of the subject. Value.

Step S35: Calculate a boundary value of the acceleration first point (Q1p, Q1p=iSize*low) and a boundary value of the acceleration to the second level (Q3p, Q3p=iSize*high) according to the array size (iSize), wherein The data of the sorted waveform value of 25% or 75% can be taken as the parameter value. Generally, the preset value is low_level=0.25, high level=0.75, and the level can be adjusted according to the user's habit to set the sensitivity, for example: high sensitivity. (low=0.2, high=0.7), medium sensitivity (low=0.3, high=0.8) and low sensitivity (low=0.4, high=0.9).

Step S36: determining whether the boundary value of the first boundary value of the acceleration and the boundary value of the acceleration value have a decimal place value. If not, step S37 is performed, and if yes, step S38 is performed.

Step S37: the value in the second quasi-point is a running step value (Q3), and the value in the first quasi-point is a walking step threshold (Q1); wherein, the Q1 and Q3 equations are as follows ( 1) and formula (2):

Step S38: Take the integer part of the first level (Q1, Q1p=take integer Q1p) and the second level point (Q3, Q3p=take integer Q3p) as the new level.

Among them, the running step value (Q3) is the average value of the value in the next position of the second position of the new leveling point, and the step threshold (Q1) of the walking is the first point of the new level. The average of the values in the previous and next positions, and the average equations of Q1 and Q3 are as shown in the following equations (3) and (4):

Step S39: storing the step threshold of the running and the step threshold of the walking for the subsequent analysis of the gait.

Since each person applies the personalized biofeedback physiological sensor of the present invention to the body part and the manner is different, the acceleration generated after the exercise is also different, thereby being known by the above personalized gait analysis method. The state of walking or running can be correctly distinguished to improve the accuracy of the measurement.

Please refer to the fourth figure, which is a flow chart of the steps of the personalized physiological sensing method of the physiological sensor of the present invention. The test subject can download a personalized information by using a wireless transmission module (such as a Bluetooth module) to match the various motion states to enable the voice module to play the corresponding music. The personalized physiological sensing method includes: Step S41: downloading a personalized information in the physiological sensor, wherein the personalized information includes a sports medical order (for example, calories burned and the heart rate should be expected), and a person preset motion information (for example, an estimated exercise path length, Information such as path coordinates or estimated exercise time) and a person's profile (height, weight, age, or ECG biometric template).

Step S42: determining whether the electrocardiogram of the subject conforms to the electrocardiogram biometric template, and if not, downloading the personalized information again, and if yes, executing step S43.

Step S43: determining whether the subject stores the parameter value of the gait analysis, wherein the parameter value of the gait analysis may be the running step value and the step threshold of the walking, if not, executing step S44, and if yes, Step S45 is performed.

Step S44: Perform a quartile algorithm to calculate a parameter value of the gait analysis.

Step S45: calculating the calorie and exercise intensity of the subject, wherein the personalized physiological sensor calculates the calorie consumption data and the exercise intensity according to the three-axis motion signal, the parameter value of the gait analysis, the electrocardiogram signal, and the GPS motion path. (such as exercise heart rate), according to the intensity of the exercise to automatically adjust the music rhythm and song selection and broadcast to the subject.

Step S46: determining whether there is a sports medical plan that achieves personalized information, if it is greater than the exercise value that should be in the exercise order, step S47 is performed, if it is less than the exercise value that should be in the exercise order, step S48 is performed, if the exercise order is met If there is a motion value in the middle, step S49 is performed.

Step S47: Output a warning voice that has exceeded the amount of exercise or load to ensure the motion of the subject.

Step S48: The amount of exercise is insufficient and the warning voice of the movement is reminded.

Step S49: output a prompt voice and play music to the subject, such as outputting the encouraged subject Prompt voice and play specific music to improve the user's enjoyment of sports and usage.

Although the present invention has been disclosed above in the foregoing embodiments, it is not intended to limit the invention. It is within the scope of the invention to be modified and modified without departing from the spirit and scope of the invention. Please refer to the attached patent application for the scope of protection defined by the present invention.

101‧‧‧Microprocessor

102‧‧‧Three-axis accelerator sensor module

103‧‧‧Electrocard signal acquisition module

104‧‧‧Satellite Positioning Module

105‧‧‧ Analog to digital signal conversion module

106‧‧‧Voice play module

107‧‧‧Battery module

108‧‧‧Storage module

109‧‧‧Display module

110‧‧‧Wireless transmission module

111‧‧‧Input module

The first figure is a schematic diagram of a physiological sensor with biofeedback provided by an embodiment of the present invention.

The second figure is a schematic diagram of the electrocardiogram (ECG) signal of the present invention.

The third figure is a flow chart of the steps of the personalized gait analysis method of the present invention.

The fourth figure is a flow chart of the steps of the personalized physiological sensing method of the physiological sensor of the present invention.

101‧‧‧Microprocessor

102‧‧‧Three-axis accelerator sensor module

103‧‧‧Electrocard signal acquisition module

104‧‧‧Satellite Positioning Module

105‧‧‧ Analog to digital signal conversion module

106‧‧‧Voice play module

107‧‧‧Battery module

108‧‧‧Storage module

109‧‧‧Display module

110‧‧‧Wireless transmission module

111‧‧‧Input module

Claims (1)

A personalized gait analysis method for a physiological sensor detects a motion signal of a personalized running and walking various motion postures by a three-axis accelerator sensing module, and obtains the motion signal by a microprocessor to four The quantile algorithm discriminates a variety of different motion states of a subject, and the personalized gait analysis method includes: (a) inputting a preset learning mode; (b) reading the learning time required by the subject; Determining whether the learning time of the subject reaches the preset learning time, if not, performing step (d), and if so, performing step (e); (d) storing the waveform values of a three-axis accelerator module in an array (Array List), and perform step (b) again; (e) sort the waveform values stored in the array from low to high increment; (f) calculate the boundary value of the acceleration based on the array size, the first quasi-point and one The boundary value of the acceleration is the second quasi-point; (g) determining whether the first quasi-point of the boundary value of the acceleration and the second-order point of the boundary value of the acceleration have a decimal value, and if not, performing step (h), and if so, Then performing step (i); (h) then the boundary value is second The value in the punctual point is the step threshold of a running, and the value in the first level of the boundary value is a walking step threshold; (i) the first level of the boundary value is taken out and the second value of the boundary value is taken. The integer part is the new level, and the boundary value of the new level is the second and the next position. The average value of the value is used as the step threshold of the running, and the boundary value of the new level is the average value of the value in the first position and the next position of the first position as the step threshold of the walking; and (j Store the step threshold for running and the step threshold for walking.