CPR feedback detection method based on pressure detection and acceleration cosine reconstruction

Abstract

The invention discloses a CPR feedback detection method based on pressure detection and acceleration cosine reconstruction, and the method specifically comprises the following steps: S1, collecting acceleration data and pressure data, and carrying out the preprocessing; s2, pressing starting and ending points are detected according to the preprocessed pressure data, the period of changing the pressure value from the minimum point to the maximum point is recorded as a pressing period, the period of changing the pressure value from the maximum point to the minimum point is recorded as a springback period, and pressing frequency is calculated; s3, respectively performing cosine reconstruction on the accelerations of the press-down period and the rebound period according to the press starting point and the press ending point to obtain a corresponding first acceleration reconstruction coefficient and a corresponding second acceleration reconstruction coefficient; s4, according to the first acceleration reconstruction coefficient and the second acceleration reconstruction coefficient, the pressing depth and the rebound depth are calculated.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a CPR feedback detection method based on pressure detection and acceleration cosine reconstruction. Background Art

[0002] At present, the compression frequency of CPR (cardiopulmonary resuscitation) is generally calculated by detecting acceleration or pressure cycle; the compression depth is mainly measured by the second integral of acceleration, that is, the velocity is obtained by integrating the acceleration, and the displacement is obtained by integrating the velocity. The displacement difference is the compression depth. If the displacement is directly calculated by the second integral of acceleration, although it has the advantages of simple implementation and fast calculation speed, it also has the problem of integral drift, especially when the acceleration interference is complex.

[0003] Dealing with integral drift is a thorny issue. Existing technologies generally adopt compensation strategies or high-pass filtering, but the accuracy is not high enough, or introduce additional acceleration sensors or additional measurements to suppress it, which increases the complexity of the CPR feedback device. Summary of the invention

[0004] In view of the technical problems in the prior art that CPR feedback is complex or not accurate enough, the present invention proposes a CPR feedback detection method based on pressure detection and acceleration cosine reconstruction. The compression frequency is calculated by periodically detecting the pressure, the acceleration is periodically intercepted according to the periodic position of the detected pressure, and then the acceleration is cosine reconstructed. Finally, the displacement is directly calculated using the time-domain integral property of Fourier transform, thereby improving the accuracy of CPR feedback.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction specifically comprises the following steps:

[0007] S1: Collect acceleration data and pressure data and perform preprocessing;

[0008] S2: Detect the start and end points of the pressing according to the preprocessed pressure data, record the period of the pressure value changing from the minimum point to the maximum point as the pressing period, record the period of the pressure value changing from the maximum point to the minimum point as the rebound period, and calculate the pressing frequency;

[0009] S3: performing cosine reconstruction on the accelerations of the pressing period and the rebound period according to the pressing start and end points, respectively, to obtain corresponding first acceleration reconstruction coefficients and second acceleration reconstruction coefficients;

[0010] S4: Calculate the pressing depth and the rebound depth according to the first acceleration reconstruction coefficient and the second acceleration reconstruction coefficient.

[0011] Preferably, in S1, the preprocessing at least includes denoising; the denoising includes smoothing denoising using a moving mean filter and / or using a low-pass digital filter.

[0012] Preferably, the S2 includes:

[0013] S2-1: Detect the start and end points of pressing according to the pre-processed pressure data;

[0014] S2-2: Calculate the pressing frequency based on the pressing start and end points.

[0015] Preferably, in S2-1, the detection of the pressing start and end points is:

[0016] p_start=p,if:press(p)<thr_al∧press(p+1)-press(p)>thr_sl

[0017] p_end=p,if:press(p)<thr_al∧press(p-1)-press(p)<thr_sl

[0018] p_peak=p,if:press(p)>thr_ah∧press(p+1)-press(p)<thr_sh (1)

[0019] In formula (1), p_start represents the starting point of pressing; p_end represents the end point of pressing; p_peak represents the maximum point of pressing; thr_al represents the detection amplitude threshold of the starting and ending points of pressing; thr_ah represents the detection amplitude threshold of the maximum point of pressing; thr_sl represents the detection slope threshold of the starting and ending points of pressing; thr_sh represents the detection slope threshold of the maximum point of pressing; press(p) represents the pressure value of the pressure sequence at the corresponding time point p; press(p-1) represents the pressure value of the pressure sequence at the corresponding time point p-1; press(p+1) represents the pressure value of the pressure sequence at the corresponding time point p+1, and ∧ represents logical AND.

[0020] Preferably, in S2-2, the calculation formula of the pressing frequency is:

[0021] cc_rate=60*fs / (p_end-p_start) (2)

[0022] In formula (2), cc_rate represents the pressing frequency; fs represents the sampling frequency; p_start represents the pressing starting point; and p_end represents the pressing end point.

[0023] Preferably, S3 includes:

[0024] S3-1: Symmetrically extend the acceleration value sequence of the pressing cycle to obtain the extended sequence:

[0025] acc_push_sym=[acc_push, flip(acc_push)] (4)

[0026] In formula (4), acc_push_sym represents the extended sequence; acc_push represents the acceleration value sequence of the pressing cycle; flip is the sequence flip function; where,

[0027] acc_push=acc(p_start:p_peak);

[0028] S3-2: Calculate the basic angular frequency of the pressing cycle:

[0029] omiga=2*π*fs / acc_push_len (5)

[0030] In formula (5), omiga represents the basic angular frequency of the pressing cycle; fs represents the sampling frequency;

[0031] acc_push_len indicates the length of the sequence after the acceleration value of the push cycle is extended;

[0032] S3-3: Calculate the cosine sequence based on the basic angular frequency of the pressed period:

[0033] rcos(k,t)=cos(k*omiga*t*Δt);k=1,2,3;t=0,1,···,acc_push_len-1(6)

[0034] In formula (6), rcos(k,t) represents the reconstructed cosine sequence; omiga represents the basic angular frequency of the pressing cycle; k represents the cosine sequence number; t represents the sampling time; Δt represents the sampling time interval;

[0035] S3-4: Calculate the square of the length of the cosine sequence:

[0036]

[0037] In formula (7), b2 represents the square of the length of the cosine sequence;

[0038] S3-5: Calculate the first acceleration reconstruction coefficient:

[0039]

[0040] In formula (8), beta(k) represents the first acceleration reconstruction coefficient corresponding to the pressing period.

[0041] Preferably, the S4 includes:

[0042] S4-1: Calculate the displacement coefficient of the pressing cycle:

[0043] S(k)=100*beta(k) / (k*omiga) 2 ,k=1,2,3 (9)

[0044] In formula (9), S(k) represents the displacement coefficient of the pressing cycle; beta(k) represents the first acceleration reconstruction coefficient corresponding to the pressing cycle; omiga represents the basic angular frequency of the pressing cycle;

[0045] S4-2: Calculate the displacement sequence of the pressing cycle:

[0046]

[0047] In formula (10), dis_push(t) represents the periodic displacement sequence; rcos(k,t) represents the reconstructed cosine sequence; acc_push_len represents the length of the sequence after the periodic acceleration value is extended;

[0048] S4-3: Calculate the pressing depth and rebound depth:

[0049] depth_push=max(dis_push)-min(dis_push)

[0050] depth_back=max(dis_back)-min(dis_back) (11)

[0051] In formula (11), depth_push represents the pressing depth; depth_back represents the rebound depth; dis_push represents the displacement sequence of the pressing period; and dis_back represents the displacement sequence of the rebound period.

[0052] Preferably, the pressing depth is calculated according to the average value of the pressing depth and the rebound depth:

[0053] depth=(depth_push+depth_back) / 2 (12)

[0054] In formula (12), depth represents the pressing depth; depth_push represents the pressing depth; and depth_back represents the rebound depth.

[0055] Preferably, whether the rebound is sufficient is judged according to the difference between the pressing depth and the rebound depth. The judgment threshold is thr_dep. When the difference between the pressing depth and the rebound depth is less than the threshold, it can be considered that the rebound is sufficient. When the difference between the pressing depth and the rebound depth is greater than the threshold, it can be considered that the rebound is insufficient, that is:

[0056]

[0057] In formula (13), depth_push represents the pressing depth; depth_back represents the rebound depth; and thr_dep represents the threshold value.

[0058] In summary, due to the adoption of the above technical solution, compared with the prior art, the present invention has at least the following beneficial effects:

[0059] To address the quadratic integration problem of CPR feedback, the present invention calculates the pressing frequency by periodically detecting the pressure without adding additional sensors, periodically intercepts the acceleration according to the periodic position of the detected pressure, reconstructs the acceleration by cosine, and finally directly calculates the displacement using the time-domain integral property of Fourier transform, thereby improving the accuracy of CPR feedback.

[0060] At the same time, it avoids directly performing a quadratic integration on the acceleration, thereby avoiding the difficult problem of dealing with integral drift; at the same time, it also avoids directly performing a Fourier transform on the acceleration, reducing the calculation and memory requirements, and the technical solution is simple to implement, has a fast calculation speed, and has a high feedback accuracy, and can be used for general CPR feedback equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 Schematic diagram of a method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction according to an exemplary embodiment of the present invention.

[0062] Figure 2 The figure is a schematic diagram of a process of performing cosine reconstruction on acceleration according to an exemplary embodiment of the present invention.

[0063] Figure 3 Schematic diagram of a calculation flow of pressing depth and rebound depth according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0064] The present invention is further described in detail below in conjunction with the examples and specific implementation methods. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following examples, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0065] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0066] like Figure 1 As shown, the present invention provides a method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction, which specifically includes the following steps:

[0067] S1: respectively collecting corresponding acceleration data and pressure data through the existing acceleration sensor and pressure sensor on the CPR feedback device, and performing preprocessing.

[0068] In this embodiment, the preprocessing includes at least denoising. To simplify the denoising, the acceleration and pressure data are smoothed and denoised using moving mean filtering, respectively; other denoising methods such as low-pass digital filters and Kalman filters may also be used, for example, a 3rd-order Butterworth low-pass filter with a cutoff frequency of 20 Hz is used, and this embodiment does not impose any specific restrictions.

[0069] S2: Detect the start and end points of the pressing according to the preprocessed pressure data, record the period when the pressure value changes from the minimum point to the maximum point as the pressing period, record the period when the pressure value changes from the maximum point to the minimum point as the rebound period, and calculate the pressing frequency.

[0070] According to actual measurements, using pressure data to detect the start and end points of pressing is more accurate and reliable than using acceleration signals to detect. Therefore, this embodiment uses pressure data to detect the start and end points of pressing.

[0071] In this embodiment, the period during which the pressure value changes from the minimum point to the maximum point is recorded as the pressing period, and the period during which the pressure value changes from the maximum point to the minimum point is recorded as the rebound period.

[0072] The specific steps include:

[0073] S2-1: Detect the start and end points of pressing according to the pre-processed pressure data.

[0074] In this embodiment, in the pre-processed pressure data, the pressing start and end points and the maximum pressing point are detected by a dual threshold method of setting an amplitude threshold and a slope threshold respectively.

[0075] For example, suppose the preprocessed pressure data sequence is press, the current pressure point position is p, the pressing starting point is p_start (pressing start time), the pressing end point is p_end (pressing end time), the maximum pressing point is p_peak (pressing to the deepest point), the amplitude thresholds are thr_al and thr_ah, and the slope thresholds are thr_sl and thr_sh. The pressing starting and ending points are detected according to the following formula:

[0076] p_start=p,if:press(p)<thr_al∧press(p+1)-press(p)>thr_sl

[0077] p_end=p,if:press(p)<thr_al∧press(p-1)-press(p)<thr_sl

[0078] p_peak=p,if:press(p)>thr_ah∧press(p+1)-press(p)<thr_sh (1)

[0079] In formula (1), p_start represents the starting point of pressing; p_end represents the end point of pressing; p_peak represents the maximum point of pressing; thr_al represents the detection amplitude threshold of the starting and ending points of pressing; thr_ah represents the detection amplitude threshold of the maximum point of pressing; thr_sl represents the detection slope threshold of the starting and ending points of pressing; thr_sh represents the detection slope threshold of the maximum point of pressing; press(p) represents the pressure value of the pressure sequence at the corresponding time point p; press(p-1) represents the pressure value of the pressure sequence at the corresponding time point p-1; press(p+1) represents the pressure value of the pressure sequence at the corresponding time point p+1, and ∧ represents logical AND.

[0080] It should be noted that the detection of the pressing start and end points may also adopt different detection methods according to actual needs, and this embodiment does not make any specific limitation.

[0081] S2-2: Calculate the pressing frequency based on the pressing start and end points.

[0082] In this embodiment, the pressing frequency is cc_rate, the unit is times / minute, the sampling frequency is fs, the unit is Hz, the pressing starting point is p_start, and the pressing end point is p_end. The pressing frequency is calculated according to the following formula:

[0083] cc_rate=60*fs / (p_end-p_start) (2)

[0084] In formula (2), cc_rate represents the pressing frequency; fs represents the sampling frequency; p_end represents the pressing end point; and p_start represents the pressing starting point.

[0085] S3: Figure 2 As shown, cosine reconstruction is performed on the acceleration of the pressing period and the rebound period respectively to obtain the corresponding first acceleration reconstruction coefficient and second acceleration reconstruction coefficient.

[0086] In this embodiment, the acceleration is reconstructed using the least square method using a plurality of cosine functions of specific frequencies that are multiples of the base frequency (pressing frequency).

[0087] In this embodiment, let the acceleration value sequence be acc, the current acceleration value position be p, record the acceleration value sequence of the pressing cycle as acc_push, record the acceleration value sequence of the rebound cycle as acc_back, the pressing starting point as p_start, the maximum pressing point as p_peak, and the pressing end point as p_end, and intercept the acceleration value sequences of a pressing cycle and a rebound cycle respectively:

[0088] acc_push=acc(p_start:p_peak), acc_back=acc(p_peak:p_end) (3)

[0089] In this embodiment, the specific steps of performing cosine reconstruction on the acceleration of the pressing cycle include (this embodiment takes the pressing cycle as an example for explanation, and the rebound cycle processing method is the same and will not be repeated):

[0090] S3-1: Symmetrically extend the acceleration value sequence of the pressing cycle to obtain an extended sequence.

[0091] In this embodiment, the acceleration value sequence after extension is assumed to be acc_push_sym, flip is a sequence flip function, and the specific extension method of the acceleration value sequence is as follows:

[0092] acc_push_sym=[acc_push, flip(acc_push)] (4)

[0093] In formula (4), acc_push_sym represents the extended sequence; acc_push represents the acceleration value sequence of the pressing cycle; flip is the sequence flip function.

[0094] S3-2: Calculate the basic angular frequency of the pressing cycle.

[0095] In this embodiment, assuming that the basic angular frequency of the pressing cycle is omiga, the length of the acceleration value sequence acc_push_sym after the pressing cycle is extended is acc_push_len, the sampling frequency is fs, and the basic angular frequency is specifically calculated as follows:

[0096] omiga=2*π*fs / acc_push_len (5)

[0097] In formula (5), omiga represents the basic angular frequency of the pressing cycle; fs represents the sampling frequency;

[0098] acc_push_len indicates the length of the sequence after the push cycle acceleration value is extended.

[0099] S3-3: Calculate the cosine sequence based on the basic angular frequency of the pressed period.

[0100] In this embodiment, three cosine functions with the same phase and multiples of the basic frequency are used to reconstruct the acceleration after extension by least squares. More cosine functions can also be used to reconstruct the acceleration, which is not specifically limited in this embodiment. Assuming the basic angular frequency is omiga, the cosine sequence is rcos, and the sampling time interval is Δt=1 / fs, the cosine sequence is specifically calculated as:

[0101] rcos(k,t)=cos(k*omiga*t*Δt);k=1,2,3;t=0,1,···,acc_push_len-1(6)

[0102] In formula (6), rcos(k, t) represents the reconstructed cosine sequence; omiga represents the basic angular frequency of the pressing cycle; k represents the cosine sequence number; t represents the sampling time; Δt represents the sampling time interval; acc_push_len represents the length of the acceleration value sequence after the pressing cycle is extended.

[0103] S3-4: Calculate the square of the length of the cosine sequence.

[0104] In this embodiment, the acceleration is reconstructed using the same-phase cosine function whose frequency is a multiple of the basic frequency. These cosine sequences have equal lengths and are orthogonal to each other; therefore, any one of them can be selected to calculate the square parameter of the cosine sequence length.

[0105] In this embodiment, assuming that the square of the length of the cosine sequence is b2, and the cosine sequence is rcos, the square of the length of the cosine sequence is specifically calculated as follows:

[0106]

[0107] In formula (7), b2 represents the square of the length of the cosine sequence; rcos(k, t) represents the reconstructed cosine sequence; acc_push_len represents the length of the acceleration value sequence after the push cycle extension.

[0108] S3-5: Calculate the first acceleration reconstruction coefficient.

[0109] In this embodiment, the first acceleration reconstruction coefficient is assumed to be beta, and the acceleration reconstruction coefficient is specifically calculated as follows:

[0110]

[0111] In formula (8), beta(k) represents the first acceleration reconstruction coefficient corresponding to the pressing cycle, where k represents the cosine sequence number; b2 represents the square of the cosine sequence length; rcos(k, t) represents the reconstructed cosine sequence; acc_push_sym represents the acceleration value sequence after the pressing cycle is extended; acc_push_len represents the length of the acceleration value sequence after the pressing cycle is extended.

[0112] Similarly, according to the above method, the calculation method of the acceleration reconstruction coefficient of the rebound period is consistent with that of the pressing period, and the second acceleration reconstruction coefficient corresponding to the rebound period can be obtained.

[0113] S4: Figure 3 As shown, the pressing depth and the rebound depth are calculated according to the first acceleration reconstruction coefficient and the second acceleration reconstruction coefficient.

[0114] According to the time domain integration property of Fourier transform, time domain integration is equivalent to frequency domain division. In this embodiment, the displacement can be directly calculated according to the first acceleration reconstruction coefficient and the second acceleration reconstruction coefficient. The specific steps of calculating the displacement and depth include:

[0115] S4-1: Calculate the displacement coefficient of the pressing cycle.

[0116] In this embodiment, assuming that the displacement coefficient is S, the acceleration reconstruction coefficient is beta, and the basic angular frequency is omiga, the displacement coefficient is calculated as follows:

[0117] S(k)=100*beta(k) / (k*omiga) 2 ,k=1,2,3 (9)

[0118] In formula (9), S(k) represents the displacement coefficient of the pressing cycle; beta(k) represents the first acceleration reconstruction coefficient corresponding to the pressing cycle; omiga represents the basic angular frequency of the pressing cycle.

[0119] Similarly, according to the above method, the displacement coefficient of the rebound period can be obtained.

[0120] S4-2: Calculate the displacement sequence of the pressing cycle.

[0121] In this embodiment, the periodic displacement sequence is assumed to be dis_push, the unit is centimeter, the displacement coefficient is S, the cosine sequence is rcos, and the periodic displacement sequence is calculated as follows:

[0122]

[0123] In formula (10), dis_push(t) represents the displacement sequence of the pressing period; S(k) represents the displacement coefficient of the pressing period; rcos(k,t) represents the reconstructed cosine sequence; and acc_push_len represents the sequence length after the acceleration value of the pressing period is extended.

[0124] Similarly, according to the above method, the displacement sequence dis_back of the rebound period can be obtained.

[0125] S4-3: Calculate the pressing depth and rebound depth.

[0126] In this embodiment, the pressing depth is depth_push, the rebound depth is depth_back, the pressing periodic displacement sequence is dis_push, the rebound periodic displacement sequence is dis_back, the maximum value function is max, and the minimum value function is min. The pressing depth and the rebound depth are calculated according to the following formula:

[0127] depth_push=max(dis_push)-min(dis_push)

[0128] depth_back=max(dis_back)-min(dis_back) (11)

[0129] In formula (11), depth_push represents the pressing depth; depth_back represents the rebound depth; dis_push represents the displacement sequence of the pressing period; and dis_back represents the displacement sequence of the rebound period.

[0130] In addition, the pressing depth can be calculated based on the average of the pressing depth and the rebound depth. Suppose the pressing depth is depth, it can be calculated as:

[0131] depth=(depth_push+depth_back) / 2 (12)

[0132] In formula (12), depth represents the pressing depth; depth_push represents the pressing depth; and depth_back represents the rebound depth.

[0133] In addition, whether the rebound is sufficient can be judged based on the difference between the pressing depth and the rebound depth. The judgment threshold is set to thr_dep. When the difference between the pressing depth and the rebound depth is less than the threshold, it can be considered that the rebound is sufficient. When the difference between the pressing depth and the rebound depth is greater than the threshold, it can be considered that the rebound is insufficient, that is:

[0134]

[0135] In formula (13), depth_push represents the pressing depth; depth_back represents the rebound depth; and thr_dep represents the threshold value.

[0136] The significance of the present invention is that, in view of the integral drift problem existing in the calculation of displacement by quadratic integration of acceleration, a new method for calculating displacement is proposed by performing cosine reconstruction on the acceleration signal based on the time-domain integral property of Fourier transform. The method makes full use of the time-domain integral property of Fourier transform, directly calculates displacement from the reconstructed acceleration, avoids directly performing quadratic integration on the acceleration, and thus avoids dealing with the thorny integral drift problem. At the same time, the method avoids directly performing fast Fourier transform (FFT) on the acceleration to calculate the spectrum information, thereby reducing the computational complexity and memory requirements of the algorithm, thereby improving the efficiency of the algorithm, while retaining the main advantages of frequency domain integration.

[0137] Those skilled in the art will appreciate that the above-mentioned embodiments are specific examples for implementing the present invention, and in actual applications, various changes may be made thereto in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction, characterized in that: The specific steps include: S1: Collect acceleration data and pressure data and perform preprocessing; S2: Detect the start and end points of the pressing according to the preprocessed pressure data, record the period of the pressure value changing from the minimum point to the maximum point as the pressing period, record the period of the pressure value changing from the maximum point to the minimum point as the rebound period, and calculate the pressing frequency; S3: performing cosine reconstruction on the accelerations of the pressing period and the rebound period according to the pressing start and end points, respectively, to obtain corresponding first acceleration reconstruction coefficients and second acceleration reconstruction coefficients; S4: Calculate the pressing depth and the rebound depth according to the first acceleration reconstruction coefficient and the second acceleration reconstruction coefficient.

2. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 1, characterized in that: In S1, the preprocessing at least includes denoising; the denoising includes smoothing denoising using moving mean filtering and / or using a low-pass digital filter.

3. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 1, characterized in that: The S2 includes: S2-1: Detect the start and end points of pressing according to the pre-processed pressure data; S2-2: Calculate the pressing frequency based on the pressing start and end points.

4. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 3, characterized in that: In S2-1, the detection of the pressing start and end points is as follows: p_start=p,if:press(p)<thr_al∧press(p+1)-press(p)>thr_sl p_end=p,if:press(p)<thr_al∧press(p-1)-press(p)<thr_sl p_peak=p,if:press(p)>thr_ah∧press(p+1)-press(p)<thr_sh (1) In formula (1), p_start represents the starting point of pressing; p_end represents the end point of pressing; p_peak represents the maximum point of pressing; thr_al represents the detection amplitude threshold of the starting and ending points of pressing; thr_ah represents the detection amplitude threshold of the maximum point of pressing; thr_sl represents the detection slope threshold of the starting and ending points of pressing; thr_sh represents the detection slope threshold of the maximum point of pressing; press(p) represents the pressure value of the pressure sequence at the corresponding time point p; press(p-1) represents the pressure value of the pressure sequence at the corresponding time point p-1; press(p+1) represents the pressure value of the pressure sequence at the corresponding time point p+1, and ∧ represents logical AND.

5. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 3, characterized in that: In S2-2, the calculation formula of the pressing frequency is: cc_rate=60*fs / (p_end-p_start) (2) In formula (2), cc_rate represents the pressing frequency; fs represents the sampling frequency; p_start represents the pressing starting point; and p_end represents the pressing end point.

6. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 1, characterized in that: The S3 includes: S3-1: Symmetrically extend the acceleration value sequence of the pressing cycle to obtain the extended sequence: acc_push_sym=[acc_push, flip(acc_push)] (4) In formula (4), acc_push_sym represents the extended sequence; acc_push represents the acceleration value sequence of the pressing cycle; flip is the sequence flip function; where, acc_push=acc(p_start:p_peak); S3-2: Calculate the basic angular frequency of the pressing cycle: omiga=2*π*fs / acc_push_len (5) In formula (5), omiga represents the basic angular frequency of the pressing cycle; fs represents the sampling frequency; acc_push_len indicates the length of the sequence after the acceleration value of the push cycle is extended; S3-3: Calculate the cosine sequence based on the basic angular frequency of the pressed period: rcos(k,t)=cos(k*omiga*t*Δt);k=1,2,3;t=0,1,···,acc_push_len-1 (6) In formula (6), rcos(k,t) represents the reconstructed cosine sequence; omiga represents the basic angular frequency of the pressing cycle; k represents the cosine sequence number; t represents the sampling time; Δt represents the sampling time interval; S3-4: Calculate the square of the length of the cosine sequence: In formula (7), b2 represents the square of the length of the cosine sequence; S3-5: Calculate the first acceleration reconstruction coefficient: In formula (8), beta(k) represents the first acceleration reconstruction coefficient corresponding to the pressing period.

7. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 1, characterized in that: The S4 includes: S4-1: Calculate the displacement coefficient of the pressing cycle: S(k)=100*beta(k) / (k*omiga) 2 ,k=1,2,3 (9) In formula (9), S(k) represents the displacement coefficient of the pressing cycle; beta(k) represents the first acceleration reconstruction coefficient corresponding to the pressing cycle; omiga represents the basic angular frequency of the pressing cycle; S4-2: Calculate the displacement sequence of the pressing cycle: In formula (10), dis_push(t) represents the periodic displacement sequence; rcos(k,t) represents the reconstructed cosine sequence; acc_push_len represents the length of the sequence after the periodic acceleration value is extended; S4-3: Calculate the pressing depth and rebound depth: depth_push=max(dis_push)-min(dis_push) depth_back=max(dis_back)-min(dis_back) (11) In formula (11), depth_push represents the pressing depth; depth_back represents the rebound depth; dis_push represents the displacement sequence of the pressing period; and dis_back represents the displacement sequence of the rebound period.

8. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 7, characterized in that: The compression depth is calculated based on the average of the compression depth and the rebound depth: depth=(depth_push+depth_back) / 2 (12) In formula (12), depth represents the pressing depth; depth_push represents the pressing depth; and depth_back represents the rebound depth.

9. A method for CPR feedback detection based on pressure detection and acceleration cosine reconstruction as claimed in claim 7, characterized in that: Whether the rebound is sufficient is determined based on the difference between the pressing depth and the rebound depth. The judgment threshold is thr_dep. When the difference between the pressing depth and the rebound depth is less than the threshold, it can be considered that the rebound is sufficient. When the difference between the pressing depth and the rebound depth is greater than the threshold, it can be considered that the rebound is insufficient, that is: In formula (13), depth_push represents the pressing depth; depth_back represents the rebound depth; and thr_dep represents the threshold value.

Images