A spectrum peak identification method, device, medium and product

Abstract

The present invention relates to a spectrum peak identification method, equipment, medium and product, the method comprising the following steps: performing continuous wavelet transform on the data to be analyzed to form a wavelet coefficient matrix; graying the wavelet coefficient matrix; using a histogram valley threshold segmentation algorithm to distinguish the peaks and valleys of the grayed wavelet coefficient matrix; using an extreme value method to find the maximum value of the peaks one by one, and the position of the maximum value is the position of the spectrum peak. Based on the continuous wavelet transform, a wavelet transform matrix grayscale image is formed, and the regional position of the peak is identified by the histogram valley threshold segmentation technology, which makes up for the problem that the traditional continuous wavelet transform misses the peak with smaller intensity.

Description

Technical Field

[0001] The present invention relates to the technical field of spectrum peak recognition, and in particular to a spectrum peak recognition method, equipment, medium and product. Background Art

[0002] Mass spectrometry has the advantages of being fast, accurate, and highly sensitive, and has been widely used in the field of bioanalysis in recent years. As an indispensable part of detection, peak recognition in mass spectrometry spectra can be used for quantitative analysis because the peaks and peak area information obtained can be used for quantitative analysis. However, when testing samples, the spectra may be disturbed for some reasons, resulting in irregular peak shapes and increased difficulty in peak recognition. Therefore, an effective peak recognition method is urgently needed to process mass spectrometry data.

[0003] In recent years, experts and scholars at home and abroad have proposed a variety of chromatographic peak identification methods. Common spectrum detection methods mainly include the first-order derivative method, the second-order derivative method, and various similar methods derived from this. However, the algorithm based on the derivative method has disadvantages in accurate peak identification. Some experts and scholars chose the Mexican hat as the wavelet mother function and proposed the ridge peak search method for the first time, which effectively improved the accuracy of mass spectrometry peak search. However, the peak recognition algorithm based on continuous wavelet transform, although the recognition is relatively accurate, is prone to miss peaks with lower intensity. Summary of the invention

[0004] In order to achieve the above-mentioned purpose and other advantages according to the present invention, the first object of the present invention is to provide a spectrum peak identification method, comprising the following steps:

[0005] Perform continuous wavelet transform on the data to be analyzed to form a wavelet coefficient matrix;

[0006] Gray-scaling the wavelet coefficient matrix;

[0007] The histogram valley threshold segmentation algorithm is used to distinguish the peaks and valleys of the grayscale wavelet coefficient matrix;

[0008] The extreme value method is used to find the maximum value of the peaks one by one, and the position of the maximum value is the position of the spectrum peak.

[0009] Furthermore, the Mexican hat wavelet is selected as the wavelet basis function of continuous wavelet transform. The formula of the Mexican hat wavelet is:

[0010]

[0011] Furthermore, the step of performing continuous wavelet transform on the data to be analyzed to form a wavelet coefficient matrix comprises the following steps:

[0012] A Gaussian function is used to simulate a single spectrum peak, and the Mexican Hat wavelet is used as the wavelet basis for continuous wavelet transform to obtain the wavelet coefficient matrix.

[0013] The scale of the wavelet and the wavelet coefficient matrix form a two-dimensional matrix of wavelet coefficients. The formula of the two-dimensional matrix of wavelet coefficients is:

[0014]

[0015] Among them, s(t) is the signal to be analyzed, a is the scale parameter, b is the displacement parameter, Ψa,b(t) is the wavelet basis function, and C(a,b) is the transformed two-dimensional wavelet coefficient matrix.

[0016] Furthermore, the graying of the wavelet coefficient matrix comprises the following steps:

[0017] The two-dimensional matrix of wavelet coefficients is grayed out, and all data are changed to data between 0 and 255.

[0018] Furthermore, the method of using the histogram valley threshold segmentation algorithm to distinguish the peaks and valleys of the grayscale wavelet coefficient matrix includes the following steps:

[0019] Calculating a histogram of the two-dimensional matrix of wavelet coefficients after grayscale conversion, and performing normalization processing on the histogram;

[0020] Scan forward from the last grayscale value peak in the histogram, and each time a minimum value is found, it is compared with the previously obtained minimum value, and the smallest minimum value point is retained to obtain the threshold segmentation point between the peak area and the trough area.

[0021] Furthermore, the method of searching for the maximum value of the peaks one by one by using the extreme value method, wherein the position of the maximum value is the position of the spectrum peak, comprises the following steps:

[0022] Traverse all the peak area positions and find the maximum value of the peak. The position of the maximum value is the position of the spectrum peak.

[0023] Furthermore, it also includes using a deep learning method to segment and identify the image of the wavelet coefficient matrix.

[0024] The second object of the present invention is to provide an electronic device, comprising: a memory on which a program code is stored; a processor connected to the memory, and when the program code is executed by the processor, a spectrum peak identification method is implemented.

[0025] A third object of the present invention is to provide a computer-readable storage medium having program instructions stored thereon, wherein the program instructions implement a spectrum peak identification method when executed.

[0026] A fourth object of the present invention is to provide a computer program product, comprising a computer program / instruction, which implements a spectrum peak identification method when executed by a processor.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The present invention provides a spectrum peak recognition method, which relies on the signal to be analyzed itself, uses continuous wavelet transform to calculate the wavelet matrix of the original data, grays the wavelet matrix, and then uses the histogram valley threshold segmentation technology to accurately distinguish the peaks and valleys of the wavelet coefficient matrix. Finally, the extreme value method is used to find the maximum value of the peaks one by one. The location of the maximum value is the location of the spectrum peak.

[0029] The present invention proposes an effective peak recognition method for mass spectrometry signals. The method forms a wavelet transform matrix grayscale image based on continuous wavelet transform, and identifies the regional position of the peak through the histogram valley threshold segmentation technology, which makes up for the problem of missing peaks with smaller intensity based on traditional continuous wavelet transform.

[0030] The peak recognition method of the present invention adopts the histogram valley threshold segmentation technology, can quickly perform peak recognition, has low algorithm complexity, and is conducive to commercial applications.

[0031] The present invention converts a two-dimensional CWT coefficient matrix into a grayscale image, separates the ridge area in the CWT coefficient matrix from the background, and eliminates the false ridges generated by noise or other interfering signals. The segmentation threshold is determined using a histogram statistical method. The present invention combines the ridge and valley information in the matrix to better identify overlapping peaks. The method results proposed by the present invention are insensitive to parameters.

[0032] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention in conjunction with the accompanying drawings. The specific implementation of the present invention is given in detail by the following embodiments and their accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0034] Figure 1 A flow chart of a spectrum peak identification method in Example 1;

[0035] Figure 2 It is a schematic diagram of the wavelet coefficient matrix;

[0036] Figure 3 It is a schematic diagram of the two-dimensional matrix of wavelet coefficients;

[0037] Figure 4 is the grayscale image of the two-dimensional matrix of wavelet coefficients;

[0038] Figure 5 is the wavelet transform matrix histogram;

[0039] Figure 6 This is a schematic diagram of an electronic device according to Embodiment 2. DETAILED DESCRIPTION

[0040] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0041] CWT is an important time-frequency analysis method that converts a continuous time domain signal into the wavelet domain to form a two-dimensional wavelet coefficient matrix. The transformation formula is as follows:

[0042]

[0043] Among them, s(t) is the signal to be analyzed, a is the scale parameter, b is the displacement parameter, Ψa,b(t) is the wavelet basis function, and C(a,b) is the transformed two-dimensional wavelet coefficient matrix.

[0044] The basic process of continuous wavelet transform is to select a wavelet basis function, match the scaling and translation of the wavelet basis with the signal to be analyzed, that is, convolution operation, to form a wavelet coefficient matrix. The coefficient size represents the similarity between the signal and the wavelet basis. The peak detection process is to find the position that meets the peak shape characteristics in the spectrum, which is consistent with the principle of wavelet transform. Select a wavelet basis function similar to the peak shape, and perform continuous wavelet transform on the original spectrum. The place with large wavelet coefficient value is the position that meets the peak shape characteristics, that is, the peak position. According to this principle, a peak position detection method based on continuous wavelet transform is formed.

[0045] Mass spectrometry peaks are generally Gaussian peaks, and the Mexican hat wavelet (mexh) has the characteristics of a unique main peak and approximate symmetry, similar to a Gaussian peak shape, and is usually selected as the wavelet basis function of the CWT peak search method. Its formula is as follows:

[0046]

[0047] Example 1

[0048] A peak identification method, such as Figure 1 As shown, the following steps are included:

[0049] S1. Perform continuous wavelet transform on the data to be analyzed, such as mass spectrum data, to form a wavelet coefficient matrix. It should be noted that the peak recognition method provided by the present invention can not only effectively recognize the peaks of mass spectrum signals, but also can effectively recognize the peaks of other signals.

[0050] Preferably, the Mexican hat wavelet is selected as the wavelet basis function of the continuous wavelet transform, and the formula of the Mexican hat wavelet is:

[0051]

[0052] In this embodiment, a Gaussian function is used to simulate a single spectrum peak, and a continuous wavelet transform is performed using the Mexican hat wavelet as the wavelet basis to obtain a wavelet coefficient matrix as shown in FIG. Figure 2 As shown. In the wavelet coefficient curve at each scale, the larger the coefficient value, the higher the similarity between the signal curve and the wavelet basis function. Wavelet coefficients of different scales are obtained by scaling the wavelet basis function and convolution of the signal, and wavelet bases of different scales have self-similarity. Therefore, the wavelet coefficients will form a local maximum at the peak position of the spectrum at each scale, thus presenting a ridge shape perpendicular to the mass-to-charge ratio (m / z) axis, which is called the ridge line. The peak position detection can be achieved by locating the ridge line position.

[0053] After continuous wavelet transform, the wavelet scale and wavelet coefficient matrix form a two-dimensional matrix of wavelet coefficients as shown in Figure 3 As shown, the two-dimensional matrix formula of wavelet coefficients is:

[0054]

[0055] Among them, s(t) is the signal to be analyzed, a is the scale parameter, b is the displacement parameter, Ψa,b(t) is the wavelet basis function, and C(a,b) is the transformed two-dimensional wavelet coefficient matrix.

[0056] Figure 3 The positions of the peak ridges and valleys can be clearly seen in the image, that is, the place with the highest brightness is where the peak is located, so the regional position of the peak can be identified through image segmentation technology.

[0057] S2. Grayscale the wavelet coefficient matrix. That is, grayscale the wavelet coefficient two-dimensional matrix, and change all data to data between 0 and 255. Figure 4 shown.

[0058] Then the histogram valley threshold segmentation algorithm is used for image segmentation and recognition. By observing the image of the wavelet transform coefficient matrix, it can be found that there are three main parts with obvious grayscale differences in the image: background part, peak part, and valley part, and the peak part is obviously brighter than other places. So when the background grayscale values ​​of the peak part and the valley part are very different, such as Figure 5 As shown, the corresponding histogram will have two peaks, and the valley between the two peaks is the threshold point for dividing the peak and valley areas.

[0059] S3, using a histogram valley threshold segmentation algorithm to distinguish the peaks and valleys of the grayscale wavelet coefficient matrix. Specifically, the following steps are included:

[0060] 1. Calculate the histogram of the two-dimensional matrix of wavelet coefficients after grayscale conversion, normalize the histogram, and normalize the data to 0-1, that is, set max = 1.0;

[0061] 2. Determine the last peak grayscale value in the histogram and record it as k;

[0062] 3. Scan forward from the last gray value peak in the histogram, and find a minimum value each time; that is, if p(k-2)>p(k), p(k+2)>p(k), then p(k) is a minimum value, go to 4; otherwise go to 5;

[0063] 4. Compare with the previously obtained minimum value and retain the smallest minimum point; that is, if p(k) < max, then max = p(k), T = k, go to 5; otherwise go to 6;

[0064] 5. k = k-2, go to 3;

[0065] 6. Obtain the valley point, which is the threshold dividing point between the peak area and the valley area. That is, record the value of k, which is the threshold.

[0066] S4. Use the extreme value method to find the maximum value of the peaks one by one, and the position of the maximum value is the position of the spectrum peak. Specifically, all the peak area positions are obtained, and all the peak area positions are traversed to find the maximum value of the peak. The position of the maximum value is the position of the spectrum peak.

[0067] The present invention can be applied to signal peak recognition. The result of using the method to perform peak recognition on mass spectrum data is that all peaks are recognized, even closely adjacent peaks are recognized, with high accuracy.

[0068] Image segmentation is performed based on the coefficient matrix formed by continuous wavelet changes, so peak recognition can be regarded as image recognition. It should be understood that other methods can also be used for peak recognition. For example, deep learning methods are used to segment and recognize images of wavelet coefficient matrices. However, the amount of calculation based on deep learning and other methods is too large, which is not conducive to real-time peak recognition and commercial applications.

[0069] Example 2

[0070] An electronic device 200, such as Figure 6As shown, it includes but is not limited to: a memory 201 on which program codes are stored; a processor 202, which is connected to the memory, and when the program codes are executed by the processor, a spectrum peak identification method is implemented. For a detailed description of the method, reference can be made to the corresponding description in the above method embodiment, which will not be repeated here.

[0071] Example 3

[0072] A computer-readable storage medium stores program instructions, and a spectrum peak identification method is implemented when the program instructions are executed. For a detailed description of the method, reference may be made to the corresponding description in the above method embodiment, and no further description is given here.

[0073] Example 4

[0074] A computer program product includes a computer program / instruction, and when the computer program / instruction is executed by a processor, a spectrum peak identification method is implemented. For a detailed description of the method, reference may be made to the corresponding description in the above method embodiment, and no further description is given here.

[0075] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0076] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0077] The above are only embodiments of this specification and are not intended to limit one or more embodiments of this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and transformations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of one or more embodiments of this specification. One or more embodiments of this specification One or more embodiments of this specification One or more embodiments of this specification One or more embodiments of this specification.

Claims

1. A spectrum peak identification method, characterized in that: The steps include: Perform continuous wavelet transform on the data to be analyzed to form a wavelet coefficient matrix; Gray-scaling the wavelet coefficient matrix; The histogram valley threshold segmentation algorithm is used to distinguish the peaks and valleys of the grayscale wavelet coefficient matrix; The maximum value is found for each peak one by one through the extreme value method, and the location of the maximum value is the location of the spectrum peak; The method of using the histogram valley threshold segmentation algorithm to distinguish the peaks and valleys of the grayscale wavelet coefficient matrix comprises the following steps: Calculating a histogram of the two-dimensional matrix of wavelet coefficients after grayscale conversion, and performing normalization processing on the histogram; Scan forward from the last gray value peak in the histogram, and compare each time a minimum value is found with the previously obtained minimum value, retain the smallest minimum value point, and obtain the threshold segmentation point between the peak area and the trough area; The method of searching for the maximum value of the peaks one by one by the extreme value method, wherein the position of the maximum value is the position of the spectrum peak, comprises the following steps: Traverse all the peak area positions and find the maximum value of the peak. The position of the maximum value is the position of the spectrum peak; It also includes using a deep learning method to segment and identify the image of the wavelet coefficient matrix.

2. A spectrum peak identification method according to claim 1, characterized in that: The Mexican hat wavelet is selected as the wavelet basis function of continuous wavelet transform. The formula of the Mexican hat wavelet is:

3. A spectrum peak identification method according to claim 2, characterized in that: The method of performing continuous wavelet transform on the data to be analyzed to form a wavelet coefficient matrix comprises the following steps: A Gaussian function is used to simulate a single spectrum peak, and the Mexican Hat wavelet is used as the wavelet basis for continuous wavelet transform to obtain the wavelet coefficient matrix. The scale of the wavelet and the wavelet coefficient matrix form a two-dimensional matrix of wavelet coefficients. The formula of the two-dimensional matrix of wavelet coefficients is: Among them, s(t) is the signal to be analyzed, a is the scale parameter, b is the displacement parameter, Ψa,b(t) is the wavelet basis function, and C(a,b) is the transformed two-dimensional wavelet coefficient matrix.

4. A spectrum peak identification method according to claim 3, characterized in that: The graying of the wavelet coefficient matrix comprises the following steps: The two-dimensional matrix of wavelet coefficients is grayed out, and all data are changed to data between 0 and 255.

5. An electronic device, characterized in that: include: A memory having program code stored therein; A processor is coupled to the memory and implements the method according to any one of claims 1 to 4 when the program code is executed by the processor.

6. A computer-readable storage medium, characterized in that: Program instructions are stored thereon, and when the program instructions are executed, the method according to any one of claims 1 to 4 is implemented.

7. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images