Sputum detection system and method based on breathing sound

Abstract

The invention discloses a sputum detection system and method based on breathing sound, and belongs to the field of sputum detection.The system comprises a breathing sound data acquisition module, a preprocessing module, a feature extraction module, a feature selection module and a classification module; wherein the breathing sound data acquisition module acquires breathing sound data of a user through acoustic acquisition equipment; the preprocessing module is used for carrying out data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph; the feature extraction module performs feature extraction on the breathing sound data graph based on a gray-level co-occurrence matrix; the feature selection module evaluates the value of the feature by using the information gain, and selects a preset number of related features; and the classification module is used for classifying audio samples of the selected related features and judging whether sputum exists or not. By processing the breathing sound data signal, the condition of sputum in the trachea can be objectively and accurately monitored in real time, and the working efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of sputum detection, and more particularly to a sputum detection system and method based on breathing sound. Background Art

[0002] One of the problems that intubated patients often face is the excessive accumulation of sputum in the trachea and endotracheal tube. If the sputum is not cleared in time, it may cause airway obstruction, hypoventilation and carbon dioxide retention, which will affect the treatment effect and even endanger the patient's life. Therefore, it is particularly important to quickly and effectively clear the sputum in the trachea, especially for patients on mechanical ventilation.

[0003] At present, the traditional method of sputum clearing mainly relies on medical staff to check whether there is sputum accumulation through lung auscultation. This process involves placing a stethoscope on the patient's chest and listening to the sound of a few breaths to determine whether there is sputum in the trachea. However, this skill is relatively difficult to master, and medical staff need professional training and rich experience to accurately use the sound information obtained during auscultation to determine whether there is sputum in the trachea. There are a series of problems in monitoring sputum accumulation by auscultation: First, this method significantly increases the workload of medical staff, requires more time and energy for monitoring and judgment, and has low work efficiency. Second, this is a subjective process that is easily disturbed by the patient's bedside noise and other environmental factors, thereby affecting the accuracy of the judgment. The subjective feelings of medical staff may be affected by individual differences and external factors, resulting in inconsistency and uncertainty in the results.

[0004] Therefore, how to monitor the condition of sputum in the trachea in real time and objectively, reduce the workload of medical staff, and improve the accuracy and practicality of sputum monitoring is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides a sputum detection system and method based on breathing sound that at least solves some of the above technical problems, which helps to realize automatic and accurate monitoring of the patient's breathing sound, reduces the dependence on the hearing level of medical personnel, simplifies the operation process, and improves ease of use. At the same time, the present invention facilitates stable and accurate identification of sputum-related sound features in complex environments, overcoming the problem that traditional methods are interfered by environmental factors.

[0006] To achieve the above object, the technical solution adopted by the present invention is:

[0007] In a first aspect, the present invention provides a sputum detection system based on breathing sound, the system comprising: a breathing sound data acquisition module, a preprocessing module, a feature extraction module, a feature selection module and a classification module, wherein:

[0008] The breathing sound data collection module is used to collect the user's breathing sound data through an acoustic collection device;

[0009] The preprocessing module is used to perform data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph;

[0010] The feature extraction module is used to extract features from the breathing sound data graph based on a gray level co-occurrence matrix;

[0011] The feature selection module uses information gain to evaluate the value of features and selects a preset number of relevant features;

[0012] The classification module is used to classify the audio sample based on the selected relevant features to determine whether sputum is present.

[0013] Optionally, in the preprocessing module, the data preprocessing process includes: data segmentation, bandpass filtering, time-frequency conversion and texture enhancement.

[0014] Optionally, in the feature extraction module, the process of extracting features from the breathing sound data graph based on the gray level co-occurrence matrix includes: calculation of energy, moment of inertia, correlation and entropy, and the calculation formula is:

[0015]

[0016] In the formula, P(i,j,d,θ) represents the gray level co-occurrence matrix element, d represents the pixel distance classification, θ represents the direction, i and j represent the gray level intensity values ​​of two pixels in the θ direction respectively; g represents the total number of gray levels, u1 represents the arithmetic mean of all pixel values ​​on the row, u2 represents the arithmetic mean of all pixel values ​​on the column, d1 represents the degree of dispersion of the row pixel value and the average value, and d2 represents the degree of dispersion of the column pixel value and the average value.

[0017] Optionally, in the feature extraction module, before performing feature extraction on the breathing sound data graph, the grayscale level of the breathing sound data graph is reduced to level 16 while maintaining a histogram shape, where the histogram shape is used to describe the distribution of texture in the image.

[0018] Optionally, the information gain is used to evaluate the value of the feature, and the calculation formula is:

[0019]

[0020] In the formula, IG represents the information gain of the feature, c k represents the category set, P(c k ) indicates that the classification result is c i The probability when P(c k|t) represents the probability when the feature includes t, P(c k |t) represents the probability when the feature includes t, m represents the total number of categories, and k represents the number of the category.

[0021] Optionally, in the classification module, a logical classifier is used to classify the audio samples based on the selected relevant features.

[0022] In a second aspect, the present invention further provides a sputum detection method based on breathing sound, which is applied to the above-mentioned sputum detection system based on breathing sound to perform sputum detection based on breathing sound, and the method comprises:

[0023] S1. Collecting the user's breathing sound data through an acoustic collection device;

[0024] S2, performing data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph;

[0025] S3, extracting features from the breathing sound data graph based on a gray level co-occurrence matrix;

[0026] S4, using information gain to evaluate the value of features and select a preset number of relevant features;

[0027] S5. Classify the audio sample based on the selected relevant features to determine whether sputum is present.

[0028] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0029] The present invention discloses a sputum detection system and method based on breathing sound. Based on the processing of breathing sound data signals, the system can monitor the sputum in the trachea in real time, objectively and accurately, which helps to reduce the workload of medical staff. The system eliminates excessive reliance on the hearing level of medical staff in sputum detection, simplifies the operation process, and improves work efficiency; at the same time, the present invention has a strong anti-interference ability, which is convenient for stably and accurately identifying sputum-related sound characteristics in complex environments, and overcomes the problem that traditional methods are interfered by environmental factors.

[0030] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0031] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0033] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0034] Figure 1 A schematic diagram of the structure of a sputum detection system based on breathing sounds provided in an embodiment of the present invention.

[0035] Figure 2 A schematic diagram of a data preprocessing process provided by an embodiment of the present invention.

[0036] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0038] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0039] See also Figure 1 As shown, an embodiment of the present invention provides a sputum detection system based on breathing sound, which can be implemented based on software MATLAB and WEKA, and mainly includes five key components: a breathing sound data acquisition module, a preprocessing module, a feature extraction module, a feature selection module and a classification module, wherein:

[0040] The breathing sound data acquisition module collects the user's breathing sound data through an acoustic acquisition device; the preprocessing module performs data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph; the feature extraction module extracts features from the breathing sound data graph based on the gray-level co-occurrence matrix; the feature selection module uses information gain to evaluate the value of features and selects a preset number of related features; the classification module classifies the audio samples based on the selected related features to determine whether sputum is present.

[0041] Combine the following Figure 1-Figure 2 As shown, the specific implementation manner and detection principle of a sputum detection system based on breathing sound of the present invention are described in detail:

[0042] In this embodiment, the patient's breathing sound data can be recorded by an acoustic acquisition device. In the data preprocessing process, see Figure 2 As shown, by detecting the silent period between breaths, the breathing sound data is segmented to obtain independent individual breathing sounds. Subsequently, each independent breathing sound is bandpass filtered and time-frequency converted using short-time Fourier transform. Next, the time-frequency data is enhanced through the texture enhancement step.

[0043] In a specific embodiment, a texture-enhanced breathing sound data graph can be obtained through data preprocessing, which is convenient for the subsequent effective extraction of the features of the breathing sound signal. In the present invention, the influence of environmental noise (such as people's conversations, etc.) is minimized through data preprocessing. In this embodiment, the main sounds captured by the acoustic acquisition device include spitting, the sound of airflow hitting the tube wall, and the working sound of the ventilator; since the high-frequency band contains some key information, special attention should be paid to the frequency band of 20 to 6000Hz. This filtering provides a clear and targeted sound signal for the subsequent processing steps. It helps to highlight the important information in the key frequency range and provides a more reliable and effective acoustic basis for sputum detection.

[0044] In feature extraction, the gray-level co-occurrence matrix (GLCM) is used to quantitatively evaluate texture parameters. The GLCM is a matrix where the number of rows and columns is equal to the gray levels G in the image. The second-order statistical probability H(i, j|d, θ) represents the relative frequency of two pixels being separated by pixel distance. There are grayscale intensity values ​​i and j in the θ direction. Where Δx and Δy represent the coordinate offsets of two adjacent pixels in the x and y directions, respectively. The matrix element P(i, j, d, θ) contains the second-order statistical probability, which can be written as:

[0045] P(i,j,d,θ)={(x,y),(x+Δx,y+Δy)|f(x,y)=i;

[0046] f(x+Δx,y+Δy)=j}

[0047] Since the gray-level co-occurrence matrix (GLCM) has a large dimension, for the GLCM estimated from the image, the grayscale needs to be reduced. Therefore, the present invention preferably reduces the grayscale of the input image to 16 levels while maintaining the histogram shape (the histogram shape is used to describe the distribution of texture in the image). This degradation process helps to simplify the calculation before the matrix calculation and ensures that the description of the image texture is still accurate enough. In GLCM, four important additional texture features need to be paid attention to, namely, angular moment (energy), moment of inertia, correlation and entropy, which are used to describe the texture of the image. The definitions of these features are as follows:

[0048] Arc second moment (energy):

[0049]

[0050] Moment of Inertia:

[0051]

[0052] Relevance:

[0053]

[0054]

[0055] entropy:

[0056]

[0057] Among them, P(i,j,d,θ) represents the gray-level co-occurrence matrix element, d represents the pixel distance classification, θ represents the direction, i and j represent the gray-level intensity values ​​of the two pixels in the θ direction respectively; g represents the total number of gray levels, u1 represents the arithmetic mean of all pixel values ​​on the row; u2 represents the arithmetic mean of all pixel values ​​on the column; d1 represents the degree of discreteness of the row pixel value and the average value, and d2 represents the degree of discreteness of the column pixel value and the average value.

[0058] In this embodiment, each feature measurement is obtained for 4 angles (θ=0°, 45°, 90° and 135°); thus 4×4=16 features are obtained.

[0059] In a specific embodiment, it is crucial to select relevant descriptive features before classification. Among the initial 16 features, some features may have less information, while others may be highly relevant redundant features. In order to determine the optimal feature set, the present invention uses the information gain selection technique. This method evaluates the value of a feature by measuring the information gain relative to the category. The information gain (IG) of a feature is defined as follows:

[0060]

[0061] In the formula, IG represents the information gain of the feature, c k represents the category set, P(c k ) indicates that the classification result is c i The probability when P(c k |t) represents the probability when the feature includes t, P(c k |t) represents the probability when the feature includes t. m represents the total number of categories, and k represents the number of the category. Preferably, the 13 most important features are selected according to the order of the values.

[0062] In the present embodiment, by classifying the audio samples of the selected features, it is possible to determine whether there is sputum. Among the many possible classification algorithms, including neural networks, Bayesian networks, regression trees and linear discriminant analysis, different algorithms will produce different effects in different application scenarios. In order to solve this problem, the present embodiment preferably divides the data into two parts: the sound with sputum and the sound without sputum, forming a binary classification problem. Logical classifier is a commonly used and efficient method to solve this type of problem, so a logical classifier is selected and implemented in the open source machine learning software WEKA. This selection is based on the good performance of the logical classifier on the binary classification problem, which is an ideal tool in the sputum detection system.

[0063] Furthermore, the embodiment of the present invention also provides a sputum detection method based on breathing sound, which is applied to a sputum detection system based on breathing sound in the above embodiment to perform sputum detection; the general working process thereof can also be referred to as follows Figure 1 As shown, the method includes:

[0064] S1. Collecting the user's breathing sound data through an acoustic collection device;

[0065] S2, performing data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph;

[0066] S3, extracting features from the breathing sound data graph based on the gray-level co-occurrence matrix;

[0067] S4, using information gain to evaluate the value of features and select a preset number of relevant features;

[0068] S5. Classify the audio sample based on the selected relevant features to determine whether sputum is present.

[0069] The embodiment of the present invention provides a sputum detection method based on breathing sounds, and its implementation principle and technical effects are the same as those of the aforementioned system embodiment. For the sake of brief description, for parts not mentioned in this embodiment, reference can be made to the corresponding contents in the aforementioned system embodiment, which will not be repeated here.

[0070] From the description of the above embodiments, those skilled in the art can know that: the present invention provides a sputum detection system and method based on breathing sound to solve the problems of subjectivity, error and cumbersome operation of traditional sputum detection methods in the treatment of intubated patients. The present invention has the following advantages: by introducing advanced sound signal processing technology, automatic objective and accurate monitoring of the patient's breathing sound is achieved, the dependence on the hearing level of medical staff is reduced, the operation process is simplified, and the ease of use and work efficiency are improved. At the same time, the present invention has a strong anti-interference ability, and can stably and accurately identify the sound characteristics related to sputum in a complex environment, overcoming the problem that the traditional method is interfered by environmental factors.

[0071] Further, refer to Figure 3 As shown, an embodiment of the present invention also provides an electronic device for sputum detection based on breathing sounds; the electronic device may include: a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may also include a computer program stored in the memory 11 and executable on the processor 10.

[0072] In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing programs or modules stored in the memory 11, and calling data stored in the memory 11.

[0073] In addition, an embodiment of the present invention also provides a storage medium on which one or more programs readable by a computing device are stored, and the one or more programs include instructions, which, when executed by the computing device, enable the computing device to execute a sputum detection method based on breathing sounds in the above embodiment.

[0074] In the embodiment of the present invention, the storage medium may be, for example, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples of storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, and any suitable combination thereof.

[0075] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as systems, methods, electronic devices or computer program products, etc. Therefore, the present invention may take the form of a complete software embodiment, a complete hardware embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0076] It should be noted that the word "comprising" does not exclude the presence of modules, components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several distinct components, and by means of a suitably programmed computer.

[0077] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0078] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A sputum detection system based on breathing sound, characterized in that: The system includes: a breathing sound data acquisition module, a preprocessing module, a feature extraction module, a feature selection module and a classification module, wherein: The breathing sound data collection module is used to collect the user's breathing sound data through an acoustic collection device; The preprocessing module is used to perform data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph; The feature extraction module is used to extract features from the breathing sound data graph based on a gray level co-occurrence matrix; The feature selection module uses information gain to evaluate the value of features and selects a preset number of relevant features; The classification module is used to classify the audio sample based on the selected relevant features to determine whether sputum is present.

2. A sputum detection system based on breathing sound according to claim 1, characterized in that: In the preprocessing module, the data preprocessing process includes: data segmentation, bandpass filtering, time-frequency conversion and texture enhancement.

3. A sputum detection system based on breathing sound according to claim 1, characterized in that: In the feature extraction module, the process of extracting features from the breathing sound data graph based on the gray-level co-occurrence matrix includes: calculation of energy, moment of inertia, correlation and entropy, and the calculation formula is: in: Where P(i,j,d,θ) represents the gray-level co-occurrence matrix element, d represents the pixel distance classification, θ represents the direction, i and j represent the gray-level intensity values ​​of the two pixels in the θ direction respectively; g represents the total number of gray levels, u1 represents the arithmetic mean of all pixel values ​​on the row; u2 represents the arithmetic mean of all pixel values ​​on the column; d1 represents the degree of discreteness of the row pixel value and the average value, and d2 represents the degree of discreteness of the column pixel value and the average value.

4. A sputum detection system based on breathing sound according to claim 3, characterized in that: In the feature extraction module, before extracting features from the breathing sound data graph, the grayscale of the breathing sound data graph is reduced to level 16 while maintaining the histogram shape, which is used to describe the distribution of texture in the image.

5. The sputum detection system based on breathing sound according to claim 1, characterized in that: In the feature selection module, the information gain is used to evaluate the value of the feature, and the calculation formula is: In the formula, IG represents the information gain of the feature, c k represents the category set, P(c k ) indicates that the classification result is c k The probability when P(c k |t) represents the probability when the feature includes t, P(c k |t) represents the probability when the feature includes t, m represents the total number of categories, and k represents the number of the category.

6. The sputum detection system based on breathing sound according to claim 1, characterized in that: In the classification module, a logical classifier is used to classify the audio samples based on the selected relevant features.

7. A sputum detection method based on breathing sound, characterized in that: A sputum detection system based on breathing sound as described in any one of claims 1 to 6 is used to perform sputum detection, the method comprising: S1. Collecting the user's breathing sound data through an acoustic collection device; S2, performing data preprocessing on the collected breathing sound data to obtain a texture-enhanced breathing sound data graph; S3, extracting features from the breathing sound data graph based on a gray level co-occurrence matrix; S4, using information gain to evaluate the value of features and select a preset number of relevant features; S5. Classify the audio sample based on the selected relevant features to determine whether sputum is present.

Images