An intelligent diagnostic system based on cardiovascular diseases

Abstract

The present invention relates to the field of medical image processing technology, and in particular to an intelligent diagnosis system based on cardiovascular diseases. The present invention sets an acquisition module and an intelligent analysis module. A first analysis unit marks a corresponding area based on special features in a regional image. A second analysis unit intercepts a time domain waveform image segment of an electrocardiogram signal and calculates a difference. When the time domain waveform image segment is in a first waveform state, a waveform characterization value is calculated based on peaks and troughs in the time domain waveform after noise reduction filtering, and it is determined whether there is a heartbeat abnormality in the time domain waveform image. When the time domain waveform image segment is in a second waveform state, noise reduction filtering is performed on any sub-waveform segment, and the segment is fitted with an abnormal sample waveform segment in a sample database to determine whether the sub-waveform segment is abnormal, thereby improving the ability to diagnose cardiovascular abnormalities.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to an intelligent diagnosis system based on cardiovascular diseases. Background Art

[0002] The total market size of cardiovascular disease (CVD) is approximately $600 billion, and one-third of American adults suffer from one or more CVDs. It is estimated that the total cost of high blood pressure (HTN) alone exceeds $90 billion, which includes the cost of medications, unnecessary complications, emergency room visits and hospitalizations. In most cases, patients with HTN have other diseases, and HTN can indicate the risk of aneurysm, heart disease, stroke, kidney failure, metabolic syndrome, heart failure and other types of CVD; therefore, CVD is a huge burden on the healthcare system. Patient follow-up (patient access) between office visits / hospitalizations is limited or non-existent, and the declining supply of general cardiologists and the increasing demand from heart patients have exacerbated this situation. Other factors lead to defects in the current methods of providing remote management of patients with CVD-related conditions. Therefore, there is a need in the field of cardiovascular disease to create new and useful methods and systems for assessing and managing cardiovascular disease.

[0003] Chinese Patent Publication No.: CN107847156A, the invention provides a method for evaluating cardiovascular disease of a user having a body region using a mobile computing device including a camera module, comprising receiving time-series image data of the body region of the user, collecting the time-series image data during a time period; generating a photoelectric vascular volume map dataset from the time-series image data; generating a processed PPG dataset; determining a cardiovascular parameter value of the user based on the processed PPG dataset; fitting a chronobiological model to (1) the cardiovascular parameter value and (2) subsequent cardiovascular parameter values, characterizing the change of the user's cardiovascular parameter over time based on the fitted chronobiological model; and presenting an analysis of the change of the cardiovascular parameter to the user at the mobile computing device.

[0004] It can be seen that the following problems still exist in the existing technology: the impact of heart morphology on cardiovascular disease is not taken into consideration, which limits the diagnosis of cardiovascular diseases and the diagnostic accuracy of the diagnostic system is low. Summary of the Invention

[0005] To overcome the problem that the prior art fails to consider the impact of heart morphology on cardiovascular disease, resulting in limitations in the diagnosis of cardiovascular disease and low diagnostic accuracy of the diagnostic system, the present invention provides an intelligent diagnostic system for cardiovascular disease, which includes:

[0006] An acquisition module, comprising an image acquisition unit for acquiring images of different regions and a plurality of sound receiving units for acquiring electrocardiogram signals of different regions;

[0007] an intelligent analysis module connected to the detection module and the image acquisition module, comprising a first analysis unit and a second analysis unit, wherein the first analysis unit is configured to receive images acquired by the image acquisition unit and mark corresponding regions based on special features in the regional images, wherein the special features include thin cardiovascular vessel walls and an enlarged heart;

[0008] The second analysis unit is connected to the first analysis unit and is used to detect the marked area, including:

[0009] controlling the sound receiving unit to collect the ECG signal within the marked area and constructing a time-domain waveform image of the ECG signal in real time, intercepting a time-domain waveform image segment at predetermined intervals, extracting a waveform profile of each sub-waveform segment in the time-domain waveform image segment, calculating a difference of the time-domain waveform image based on the waveform profile of each sub-waveform segment, and determining a state of the time-domain waveform image segment based on the difference, wherein the waveform profile includes an area and a period;

[0010] When the time-domain waveform image segment is in a first waveform state, performing noise reduction and filtering on the time-domain waveform image segment, calculating a waveform characterization value based on peaks and troughs in the time-domain waveform after noise reduction and filtering, determining whether there is a heartbeat abnormality in the time-domain waveform image based on the waveform characterization value, and determining whether there is an abnormality in the time-domain waveform image segment based on the heartbeat abnormality;

[0011] When the time-domain waveform image segment is in the second waveform state, any sub-waveform segment in the time-domain waveform image segment is extracted, noise reduction and filtering are performed on the sub-waveform segment, the sub-waveform segment is fitted with a normal sample waveform segment in a sample database, and whether the sub-waveform segment is abnormal is determined based on the fitting coincidence degree, and the determination result of the sub-waveform segment is used as the determination result of the time-domain waveform image segment.

[0012] Furthermore, the first analysis unit marks the corresponding area based on the special features in the area image, wherein:

[0013] If the cardiovascular system in the region image has thin blood vessel walls and an enlarged heart, the first analyzing unit determines to mark the corresponding region.

[0014] Furthermore, the second analysis unit is further configured to calculate an area average based on the area of each of the sub-waveform segments and to calculate a period average based on the period of each of the sub-waveform segments;

[0015] An area reference value of the sub-waveform segment is determined based on the area average value, and a period reference value of the sub-waveform segment is determined based on the period average value.

[0016] Furthermore, the second analysis unit calculates the difference of the time domain waveform image based on the waveform profile of each of the sub-waveform segments according to formula (1),

[0017]

[0018] In formula (1), E1 is the difference, A i is the area of the ith sub-waveform segment, T i is the period of the i-th sub-waveform segment, n is the number of sub-waveform segments, A0 is the area reference value of the sub-waveform segment, and T0 is the period reference value of the sub-waveform segment.

[0019] Furthermore, the second analyzing unit determines the state of the time domain waveform image segment based on the difference, wherein:

[0020] Compare the difference with a preset difference threshold,

[0021] If the difference is greater than the difference threshold, the second analyzing unit determines that the state of the time-domain waveform image segment is a first waveform state;

[0022] If the difference is less than or equal to the difference threshold, the second analyzing unit determines that the state of the time-domain waveform image segment is a second waveform state.

[0023] Furthermore, when the state of the time domain waveform image segment is the first waveform state, the second analysis unit calculates the waveform characterization value according to formula (2) based on the time domain waveform after noise reduction filtering.

[0024]

[0025] In formula (2), E2 is the waveform characterization value, m is the variance value of several peaks of the time domain waveform after noise reduction filtering, m0 is the peak variance reference value, n is the variance value of several troughs of the time domain waveform after noise reduction filtering, and n0 is the trough variance reference value.

[0026] Furthermore, the second analyzing unit determines whether there is a heartbeat abnormality in the time domain waveform image based on the waveform characterization value, wherein:

[0027] Comparing the waveform characterization value with a preset waveform characterization value reference value,

[0028] If the waveform characterization value is less than or equal to the waveform characterization value reference value, the second analyzing unit determines that there is no heartbeat abnormality in the time domain waveform image;

[0029] If the waveform characterization value is greater than the waveform characterization value reference value, the second analyzing unit determines that there is a heartbeat abnormality in the time domain waveform image.

[0030] Furthermore, the second analyzing unit determines whether the time domain waveform image segment is abnormal based on the heartbeat abnormality, wherein:

[0031] The second analyzing unit determines whether an abnormality exists in the time-domain waveform image segment when it is determined that an abnormal heartbeat exists in the time-domain waveform image.

[0032] Furthermore, the second analyzing unit determines whether the sub-waveform segment has an abnormality based on the fitting coincidence when the state of the time domain waveform image segment is the second waveform state, wherein:

[0033] Compare the fitting coincidence with a preset fitting coincidence threshold,

[0034] If the fitting coincidence is greater than the fitting coincidence threshold, the second analyzing unit determines that there is no abnormality in the sub-waveform segment;

[0035] If the fitting coincidence is less than or equal to the fitting coincidence threshold, the second analysis unit determines that an abnormality exists in the sub-waveform segment.

[0036] Furthermore, the intelligent analysis module is also connected to an alarm unit, and the alarm unit issues an alarm based on the abnormal results determined by the intelligent analysis module.

[0037] Compared with the prior art, the present invention sets up an acquisition module and an intelligent analysis module, wherein the first analysis unit marks the corresponding area based on the special features in the regional image acquired by the acquisition module, and the second analysis unit constructs a time domain waveform image of the electrocardiogram signal based on the audio acquired by the sound receiving unit, intercepts the time domain waveform image segment, calculates the difference of the time domain waveform image based on the waveform contour of the time domain waveform image segment, and determines the state of the time domain waveform image segment based on the difference. When the time domain waveform image segment is in the first waveform state, the waveform characterization value is calculated based on the peaks and troughs in the time domain waveform after noise reduction filtering, and determines whether there is a heartbeat abnormality in the time domain waveform image based on the waveform characterization value. When the time domain waveform image segment is in the second waveform state, noise reduction filtering is performed on any sub-waveform segment in the time domain waveform image segment, and the sub-waveform segment is fitted with the abnormal sample waveform segment in the sample database to determine whether the sub-waveform segment is abnormal, thereby improving the ability to diagnose cardiovascular abnormalities.

[0038] In particular, the present invention uses the first analysis unit to mark areas with abnormal cardiovascular flow rates based on the images captured by the image acquisition unit. In actual situations, if the flow rate is too high when the blood vessels in the cardiovascular system pump blood in or out, there may be safety hazards. When the blood flow rate is too high, the heart's beating frequency may also be abnormal, and the generated electrocardiogram signal may be more prominent. The electrocardiogram signal in the marked area may be more prominent and more data representative, which is convenient for subsequent corresponding analysis and processing, thereby improving the ability of the cardiovascular system to screen abnormal areas.

[0039] In particular, the present invention constructs a time domain waveform image of the electrocardiogram signal in real time through the second analysis unit, intercepts a time domain waveform image segment at predetermined intervals, and determines the status of the time domain waveform image segment. In actual situations, the interception interval of the acoustic wave time domain waveform image is set according to production monitoring requirements, and the similarity of each sub-waveform segment is obtained through a comprehensive evaluation calculation of the area and period of several sub-waveform segments in the intercepted time domain waveform image segment. The area and period frequency of the waveform graph are basic parameters of the time domain waveform image, which can characterize the similarity of each sub-waveform segment to a certain extent. Moreover, the extraction of the above basic parameters takes up less computing power, and the difference between each sub-waveform segment of the intercepted time domain waveform image is calculated more scientifically and effectively.

[0040] In particular, the present invention uses the second analysis unit to characterize the poor stability of the intercepted time domain waveform image segment and the large number of interference features when the calculated similarities of the sub-waveform segments are poor. However, since there may be external noise interfering with the waveform detection results, it is necessary to further perform noise reduction filtering on the time domain waveform image segment, and calculate the waveform characterization value of the processed waveform segment in combination with the peaks and troughs. The comprehensive calculation of the peaks and troughs can derive the stability of the time domain waveform image segment. Furthermore, by analyzing and judging the filtered time domain waveform image segment, more accurate self-diagnosis can be performed.

[0041] In particular, the present invention uses the second analysis unit to characterize the good stability and certain similarity of the intercepted time domain waveform image segment when the similarity of each sub-waveform segment calculated is good. Therefore, the sub-waveform segment can be selected for noise reduction filtering detection. The detection result of the entire intercepted time domain waveform image segment can be characterized by the local detection result. The sub-waveform segment after noise reduction filtering is fitted with the normal sample waveform of the database to determine whether the intercepted time domain waveform image segment is abnormal. The above process reduces the computing power consumption on interference noise reduction filtering, and the monitoring results are highly reliable, which improves the ability to diagnose cardiovascular diseases. It is suitable for multi-cardiovascular joint monitoring and reduces computing losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1This is a block diagram of a smart diagnostic system for cardiovascular diseases according to an embodiment of the invention;

[0043] Figure 2 This is a logic flow chart of the first analysis unit according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below with reference to embodiments. It should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0045] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0046] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside", and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0047] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediary; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0048] See also Figure 1 As shown in FIG, which is a block diagram of a module of an intelligent diagnosis system for cardiovascular diseases according to an embodiment of the invention, the intelligent diagnosis system for cardiovascular diseases according to an embodiment of the invention comprises:

[0049] An acquisition module, comprising an image acquisition unit for acquiring images of different areas and a plurality of sound receiving units for acquiring electrocardiogram signals of different areas within the cabinet;

[0050] an intelligent analysis module connected to the detection module and the image acquisition module, comprising a first analysis unit and a second analysis unit, wherein the first analysis unit is configured to receive images acquired by the image acquisition unit and mark corresponding regions based on special features in the regional images, wherein the special features include thin cardiovascular vessel walls and an enlarged heart;

[0051] The second analysis unit is connected to the first analysis unit and is used to detect the marked area, including:

[0052] controlling the sound receiving unit to collect the ECG signal within the marked area and constructing a time-domain waveform image of the ECG signal in real time, intercepting a time-domain waveform image segment at predetermined intervals, extracting a waveform profile of each sub-waveform segment in the time-domain waveform image segment, calculating a difference of the time-domain waveform image based on the waveform profile of each sub-waveform segment, and determining a state of the time-domain waveform image segment based on the difference, wherein the waveform profile includes an area and a period;

[0053] When the time-domain waveform image segment is in a first waveform state, performing noise reduction and filtering on the time-domain waveform image segment, calculating a waveform characterization value based on peaks and troughs in the time-domain waveform after noise reduction and filtering, determining whether there is a heartbeat abnormality in the time-domain waveform image based on the waveform characterization value, and determining whether there is an abnormality in the time-domain waveform image segment based on the heartbeat abnormality;

[0054] When the time-domain waveform image segment is in the second waveform state, any sub-waveform segment in the time-domain waveform image segment is extracted, noise reduction and filtering are performed on the sub-waveform segment, the sub-waveform segment is fitted with a normal sample waveform segment in a sample database, and whether the sub-waveform segment is abnormal is determined based on the fitting coincidence degree, and the determination result of the sub-waveform segment is used as the determination result of the time-domain waveform image segment.

[0055] Specifically, the present invention does not limit the specific structure of the image acquisition unit. Preferably, it can be a high-definition industrial CCD camera, which has been widely used in the field of medical visual imaging and will not be described in detail here.

[0056] Specifically, the present invention does not limit the specific structure of the sound receiving unit, and it only needs to be able to collect the working sounds inside the cardiovascular system. This is existing technology and will not be described in detail here.

[0057] Specifically, the present invention does not limit the specific structure of the intelligent analysis module and the functional units inside it. It can be a micro-control computer that can receive, process and send information data, or it can be a CPU unit that integrates relevant functional algorithms. This is existing technology and will not be repeated here.

[0058] Specifically, the present invention does not limit the specific method of noise reduction and filtering, and can use a variety of noise reduction and filtering methods that meet the accuracy to process ECG signals. In this embodiment, the wavelet method can be used to decompose the ECG signal. In the wavelet method, the ECG signal needs to be decomposed based on the decomposition scale. The data processing of this process is relatively complicated, and if higher accuracy is to be obtained, the decomposition scale needs to be increased. The increase in the decomposition scale will also be accompanied by an increase in the amount of calculation and calculation time. This is an existing technology and will not be repeated here.

[0059] Specifically, in the present invention, the time interval for intercepting the time domain waveform image segment can be set according to production requirements. The set time interval can meet the monitoring needs and can effectively reduce the computational complexity of the noise reduction filter. Those skilled in the art can set the time interval within [1,5] according to this setting logic, and the interval unit is h.

[0060] Specifically, the first analysis unit marks the corresponding area based on the special features in the area image, wherein:

[0061] If the cardiovascular system in the region image has thin blood vessel walls and an enlarged heart, the first analyzing unit determines to mark the corresponding region.

[0062] Specifically, see Figure 2 As shown, it is a logic flow chart of the first analysis unit of an embodiment of the invention. The present invention marks the area with abnormal cardiovascular flow rate based on the image collected by the image acquisition unit by the first analysis unit. In actual situations, if the flow rate is too high during the process of pumping blood into or out of the cardiovascular blood vessels, there may be safety hazards. When the blood flow rate is too high, the heart beat frequency will also be abnormal, and the generated electrocardiogram signal will be more prominent. The electrocardiogram signal in the marked area is more prominent and has more data representation, which is convenient for subsequent corresponding analysis and processing, thereby improving the ability of cardiovascular screening abnormal areas.

[0063] Specifically, the second analysis unit is further configured to calculate an area average based on the area of each of the sub-waveform segments and to calculate a period average based on the period of each of the sub-waveform segments;

[0064] and determining an area reference value A0 of the sub-waveform segment based on the area average value Aa and determining a period reference value T0 of the sub-waveform segment based on the period average value Ta;

[0065] Among them, A0=[0.1Aa,0.2Aa], T0=[0.05Ta,0.15Ta].

[0066] Specifically, the second analysis unit calculates the difference of the time domain waveform image based on the waveform profile of each sub-waveform segment according to formula (1),

[0067]

[0068] In formula (1), E1 is the difference, A i is the area of the ith sub-waveform segment, T i is the period of the i-th sub-waveform segment, n is the number of sub-waveform segments, A0 is the area reference value of the sub-waveform segment, and T0 is the period reference value of the sub-waveform segment.

[0069] Specifically, the present invention constructs a time domain waveform image of the electrocardiogram signal in real time through a second analysis unit, intercepts a time domain waveform image segment at predetermined intervals, and determines the status of the time domain waveform image segment. In actual situations, the interception interval of the acoustic wave time domain waveform image is set according to production monitoring requirements, and several sub-waveform segments in the intercepted time domain waveform image segment are comprehensively evaluated and calculated on the area and period to obtain the similarity of each sub-waveform segment. The area and period frequency of the waveform graph are basic parameters of the time domain waveform image, which can characterize the similarity of each sub-waveform segment to a certain extent. Moreover, the extraction of the above basic parameters takes up less computing power, and the difference between each sub-waveform segment of the intercepted time domain waveform image is calculated more scientifically and effectively.

[0070] Specifically, the second analyzing unit determines the state of the time domain waveform image segment based on the difference, wherein:

[0071] Compare the difference E1 with the preset difference threshold Ea,

[0072] If the difference degree E1 is greater than the difference threshold Ea, the second analyzing unit determines that the state of the time domain waveform image segment is the first waveform state;

[0073] If the difference degree E1 is less than or equal to the difference threshold Ea, the second analyzing unit determines that the state of the time domain waveform image segment is a second waveform state;

[0074] Wherein, the difference threshold Ea∈[1.5,2.5].

[0075] Specifically, when the state of the time domain waveform image segment is the first waveform state, the second analysis unit calculates the waveform characterization value according to formula (2) based on the time domain waveform after noise reduction filtering:

[0076]

[0077] In formula (2), E2 is the waveform characterization value, m is the peak variance value of the time domain waveform after noise reduction filtering, m0 is the peak variance reference value, n is the trough variance value of the time domain waveform after noise reduction filtering, and n0 is the trough variance reference value;

[0078] Among them, the peak variance reference value m0 and the trough variance reference value n0 are obtained based on measurements, and the peak variance average value and the trough variance average value of the time domain waveform image segment under normal operating conditions are obtained for several times. The measured peak variance average value is used as the peak variance reference value m0, and the measured trough variance average value is used as the trough variance reference value n0.

[0079] Specifically, the second analyzing unit determines whether there is a heartbeat abnormality in the time domain waveform image based on the waveform characterization value, wherein:

[0080] Compare the waveform characterization value E2 with the preset waveform characterization value reference value Eb,

[0081] If the waveform characterization value E2 is less than or equal to the waveform characterization value reference value Eb, the second analysis unit determines that there is no heartbeat abnormality in the time domain waveform image;

[0082] If the waveform characterization value E2 is greater than the waveform characterization value reference value Eb, the second analysis unit determines that there is a heartbeat abnormality in the time domain waveform image;

[0083] The waveform characterization value reference value Eb∈[2.2,2.8].

[0084] Specifically, the second analyzing unit determines whether the time domain waveform image segment is abnormal based on the heartbeat abnormality, wherein:

[0085] The second analyzing unit determines whether an abnormality exists in the time-domain waveform image segment when it is determined that an abnormal heartbeat exists in the time-domain waveform image.

[0086] Specifically, the present invention uses the second analysis unit to characterize the poor stability of the intercepted time domain waveform image segment and the large number of interference features when the calculated similarities of the sub-waveform segments are poor. However, since there may be external noise interfering with the waveform detection results, it is necessary to further perform noise reduction filtering on the time domain waveform image segment, and calculate the waveform characterization value of the processed waveform segment in combination with the peaks and troughs. The comprehensive calculation of the peaks and troughs can derive the stability of the time domain waveform image segment. Furthermore, by analyzing and judging the filtered time domain waveform image segment, a more accurate diagnosis can be performed.

[0087] Specifically, the second analyzing unit determines whether the sub-waveform segment has an abnormality based on the fitting coincidence when the state of the time domain waveform image segment is the second waveform state, wherein:

[0088] Compare the fitting coincidence S with the preset fitting coincidence threshold S0,

[0089] If the fitting coincidence S is greater than the fitting coincidence threshold S0, the second analyzing unit determines that there is no abnormality in the sub-waveform segment;

[0090] If the fitting coincidence is less than or equal to the fitting coincidence threshold, the second analyzing unit determines that an abnormality exists in the sub-waveform segment;

[0091] The fitting coincidence threshold S0 is obtained based on a measurement result, and the fitting average Sa of the sub-waveform segments of the time domain waveform image segment under normal operating conditions and the normal sample waveform image in the database is obtained, and the fitting average Sa is used as the fitting coincidence threshold S0;

[0092] The present invention does not specifically limit the waveform fitting method. There are many waveform fitting methods in the prior art. As fitting tools, matlab and python related fitting databases can be used, which will not be described here.

[0093] The fitting coincidence threshold S0 can be obtained by pre-statistics in the present invention, wherein the time domain waveform images corresponding to several electrocardiogram signals under abnormal cardiovascular conditions are collected and fitted with the normal sample waveform images in the database to solve the fitting average value Sa, and set S0 = γ × Sa, where γ represents the accuracy coefficient, 0.8 < γ < 1.2.

[0094] Specifically, the present invention uses the second analysis unit to characterize the good stability and certain similarity of the intercepted time domain waveform image segment when the similarity of each sub-waveform segment calculated is good. Therefore, the sub-waveform segment can be selected for noise reduction filtering detection. The detection result of the entire intercepted time domain waveform image segment can be characterized by the local detection result. The sub-waveform segment after noise reduction filtering is fitted with the normal sample waveform of the database to determine whether the intercepted time domain waveform image segment is abnormal. The above process reduces the computing power consumption on interference noise reduction filtering, and the monitoring results are highly reliable, which improves the cardiovascular self-diagnosis capability, is suitable for multi-cardiovascular joint monitoring, and reduces computing loss.

[0095] Specifically, the intelligent analysis module is also connected to an alarm unit, and the alarm unit issues an alarm prompt based on the abnormal results determined by the intelligent analysis module. In this embodiment, the alarm unit may be a sound-emitting device that issues an alarm voice prompt in response.

[0096] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. An intelligent diagnostic system for cardiovascular diseases, characterized in that: include: The acquisition module includes an image acquisition unit for acquiring images of different areas and a plurality of sound receiving units for acquiring sounds of different areas; an intelligent analysis module connected to the acquisition module, comprising a first analysis unit and a second analysis unit, wherein the first analysis unit is configured to receive the image acquired by the image acquisition unit and mark corresponding regions based on special features in the regional image, wherein the special features include thin cardiovascular vessel walls and an enlarged heart; The second analysis unit is connected to the first analysis unit and is used to detect the marked area, including: Controlling the sound receiving unit to collect sound within the marked area, the second analysis unit constructing a time-domain waveform image of the electrocardiogram signal based on the audio collected by the sound receiving unit, intercepting a time-domain waveform image segment at predetermined intervals, and extracting a waveform profile of each sub-waveform segment in the time-domain waveform image segment, calculating a difference of the time-domain waveform image based on the waveform profile of each sub-waveform segment, and determining a state of the time-domain waveform image segment based on the difference, wherein the waveform profile includes an area and a period; The second analyzing unit determines the state of the time domain waveform image segment based on the difference, wherein: Compare the difference with a preset difference threshold, If the difference is greater than the difference threshold, the second analyzing unit determines that the state of the time-domain waveform image segment is a first waveform state; If the difference is less than or equal to the difference threshold, the second analyzing unit determines that the state of the time domain waveform image segment is a second waveform state; When the time-domain waveform image segment is in a first waveform state, performing noise reduction and filtering on the time-domain waveform image segment, calculating a waveform characterization value based on peaks and troughs in the time-domain waveform after noise reduction and filtering, determining whether there is a heartbeat abnormality in the time-domain waveform image based on the waveform characterization value, and determining whether there is an abnormality in the time-domain waveform image segment based on the heartbeat abnormality; When the time-domain waveform image segment is in the second waveform state, extract any sub-waveform segment from the time-domain waveform image segment, perform noise reduction and filtering on the sub-waveform segment, fit the sub-waveform segment with a normal sample waveform segment in a sample database, determine whether the sub-waveform segment is abnormal based on a fitting coincidence, and use the determination result of the sub-waveform segment as the determination result of the time-domain waveform image segment; The second analysis unit calculates the difference of the time domain waveform image based on the waveform profile of each sub-waveform segment according to formula (1), In formula (1), E1 is the difference, A i is the area of the ith sub-waveform segment, T i is the period of the i-th sub-waveform segment, n is the number of sub-waveform segments, A0 is the area reference value of the sub-waveform segment, and T0 is the period reference value of the sub-waveform segment; The second analysis unit calculates the waveform characterization value according to formula (2) based on the time domain waveform after noise reduction filtering when the state of the time domain waveform image segment is the first waveform state. In formula (2), E2 is the waveform characterization value, m is the variance value of several peaks of the time domain waveform after noise reduction filtering, m0 is the peak variance reference value, n is the variance value of several troughs of the time domain waveform after noise reduction filtering, and n0 is the trough variance reference value.

2. The intelligent diagnostic system for cardiovascular diseases according to claim 1, characterized in that: The first analysis unit marks the corresponding area based on the special features in the area image, wherein If the cardiovascular system in the region image has thin blood vessel walls and an enlarged heart, the first analyzing unit determines to mark the corresponding region.

3. The intelligent diagnostic system for cardiovascular diseases according to claim 2, characterized in that: The second analyzing unit is further configured to calculate an area average based on the area of each of the sub-waveform segments and to calculate a period average based on the period of each of the sub-waveform segments; An area reference value of the sub-waveform segment is determined based on the area average value, and a period reference value of the sub-waveform segment is determined based on the period average value.

4. The intelligent diagnostic system for cardiovascular diseases according to claim 1, characterized in that: The second analyzing unit determines whether there is a heartbeat abnormality in the time domain waveform image based on the waveform characterization value, wherein: Comparing the waveform characterization value with a preset waveform characterization value reference value, If the waveform characterization value is less than or equal to the waveform characterization value reference value, the second analyzing unit determines that there is no heartbeat abnormality in the time domain waveform image; If the waveform characterization value is greater than the waveform characterization value reference value, the second analyzing unit determines that there is a heartbeat abnormality in the time domain waveform image.

5. The intelligent diagnostic system for cardiovascular diseases according to claim 4, characterized in that: The second analyzing unit determines whether the time domain waveform image segment has an abnormality based on the heartbeat abnormality, wherein: The second analyzing unit determines whether an abnormality exists in the time-domain waveform image segment when it is determined that an abnormal heartbeat exists in the time-domain waveform image.

6. The intelligent diagnostic system for cardiovascular diseases according to claim 5, characterized in that: The second analyzing unit determines whether the sub-waveform segment has an abnormality based on the fitting coincidence when the state of the time domain waveform image segment is the second waveform state, wherein the fitting coincidence is compared with a preset fitting coincidence threshold value, If the fitting coincidence is greater than the fitting coincidence threshold, the second analyzing unit determines that there is no abnormality in the sub-waveform segment; If the fitting coincidence is less than or equal to the fitting coincidence threshold, the second analysis unit determines that an abnormality exists in the sub-waveform segment.

7. The intelligent diagnosis system based on cardiovascular disease according to claim 6, characterized in that: The intelligent analysis module is also connected to an alarm unit, and the alarm unit issues an alarm prompt based on the abnormal results determined by the intelligent analysis module.

Images