An artificial intelligence-based breast cancer ultrasound image analysis system
Through artificial intelligence technology, multi-dimensional data acquisition and analysis are adopted to solve the problem that the existing breast cancer ultrasound image analysis system is not comprehensive enough in image acquisition and the analysis results are not accurate enough, and more accurate breast cancer diagnosis and treatment strategies are achieved.
Patent Information
- Application Number
- CN202411175788.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-08-26
AI Technical Summary
The existing breast cancer ultrasound image analysis system is limited to the basic morphology and boundaries of the lesions in image acquisition, ignoring the diversity of image data, resulting in limitations in feature extraction and classification, which affects the accurate diagnosis and treatment strategy of breast cancer. .
Using an artificial intelligence-based breast cancer ultrasound image analysis system, the data batch division module is divided into different batches according to equal time, and multi-dimensional data is collected including tumor morphological boundary characteristics, internal echo, blood flow signals, imaging and The contrast data are analyzed through the comprehensive data evaluation module to generate a comprehensive breast cancer assessment index to support diagnosis.
Improve the pertinence and accuracy of image analysis, provide multi-dimensional data acquisition and analysis, reduce the one-sidedness of the data, provide doctors with a more comprehensive and objective diagnostic basis, and timely display analysis results and early warnings through real-time feedback module. Signal.
Smart Images

Figure CN118692635B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of breast cancer ultrasound image analysis, and more specifically, to a breast cancer ultrasound image analysis system based on artificial intelligence. Background Art
[0002] With the continuous advancement of medical technology, breast cancer ultrasound image analysis systems play an increasingly important role in the early screening and diagnosis of breast cancer. As a perfect combination of medical imaging technology and artificial intelligence, the system's management and analysis capabilities for ultrasound images are particularly important. Whether it is lesion detection, morphological analysis or benign and malignant discrimination, higher requirements are placed on the real-time, clarity and intelligent processing of ultrasound images.
[0003] The existing breast cancer ultrasound image analysis system mainly includes: image acquisition step, image preprocessing step, feature extraction and classification step and human-computer interaction step. Among them, the image acquisition step can capture high-quality breast ultrasound images in real time to ensure the clarity and accuracy of the image; the image preprocessing step is responsible for denoising and enhancing the acquired images to improve the image quality; the feature extraction and classification step uses deep learning, image processing technology and other means to perform a detailed analysis of the preprocessed images, extract lesion features and perform benign and malignant judgment; the human-computer interaction step provides users with intuitive and convenient diagnostic result query, report generation and early warning notification functions.
[0004] However, in the current breast cancer ultrasound image analysis system, image acquisition is often limited to the basic morphology and boundaries of the lesion, while ignoring the diversity of image data. This single image acquisition method is difficult to fully reflect the complex situation of the lesion, resulting in limitations in subsequent feature extraction and classification, affecting the accurate diagnosis of breast cancer and the formulation of treatment strategies.
[0005] Therefore, there is an urgent need for an artificial intelligence-based breast cancer ultrasound image analysis system to solve the problems of incomplete image acquisition and inaccurate analysis results in existing systems. Summary of the invention
[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a breast cancer ultrasound image analysis system based on artificial intelligence, which solves the problems raised in the above-mentioned background technology through the following scheme.
[0007] To achieve the above object, the present invention provides the following technical solution: a breast cancer ultrasound image analysis system based on artificial intelligence, comprising:
[0008] Data batch division module: used to determine the data to be collected as target data, divide the target data into different batches according to equal time division, and mark them as 1, 2, ..., n in sequence;
[0009] Breast cancer ultrasound image data acquisition module: including a diagnostic data acquisition unit and an ultrasound data acquisition unit, wherein the diagnostic data acquisition unit is used to acquire tumor morphology boundary feature data, internal echo data, and blood flow signal data;
[0010] Breast cancer ultrasound image data analysis module: includes a diagnostic data analysis unit and an ultrasound data analysis unit, which is used to analyze the data transmitted by the breast cancer ultrasound image data acquisition module and transmit the analysis results to the comprehensive data evaluation module; the diagnostic data analysis unit includes a tumor morphology boundary feature data analysis node, an internal echo data analysis node, and a blood flow signal data analysis node; the ultrasound data analysis unit includes an imaging data analysis node and an angiography data analysis node;
[0011] Comprehensive data evaluation module: including a breast cancer ultrasound image data evaluation unit, which comprehensively analyzes the data transmitted by the breast cancer ultrasound image data analysis module and transmits the analysis results to the real-time feedback module;
[0012] Real-time feedback module: used to establish preset values of the comprehensive evaluation index of breast cancer, judge the comprehensive evaluation index value of breast cancer according to the preset values of the comprehensive evaluation index of breast cancer, generate a diagnosis report according to the judgment result and save the data in the database.
[0013] Preferably, the diagnostic data analysis unit includes a tumor morphology boundary feature data analysis node, an internal echo data analysis node, and a blood flow signal data analysis node; the ultrasound data analysis unit includes an imaging data analysis node and an angiography data analysis node.
[0014] Preferably, the tumor morphology boundary feature data include the aspect ratio Ar, the burr sign occurrence rate Mo and the boundary clarity value Bc; the internal echo data include the grayscale standard deviation Gd, the number of calcification foci Cn and the degree of posterior echo attenuation Ad; the blood flow signal data include the systolic peak flow velocity Sv and the blood flow grade Tm.
[0015] Preferably, the imaging data include spatial resolution Sr, shear wave velocity Sw and maximum elastic modulus Em; the angiography data include peak time Pt, area under the curve Ac and enhanced intensity ratio Er.
[0016] Preferably, the imaging data include spatial resolution Sr, shear wave velocity Sw and maximum elastic modulus Em; the angiography data include peak time Pt, area under the curve Ac and enhanced intensity ratio Er.
[0017] Preferably, the tumor morphology boundary feature data analysis node is used to establish a tumor morphology boundary feature data calculation model, and the tumor morphology boundary feature data transmitted by the breast cancer ultrasound image data acquisition module is imported into the tumor morphology boundary feature data calculation model to obtain a tumor morphology feature value; the tumor morphology boundary feature data calculation model is specifically expressed as:
[0018] ,
[0019] in, Ar represents the tumor morphological characteristic value calculated for the i-th time; i Ar represents the aspect ratio of the i-th acquisition; i-1 Mo represents the aspect ratio of the i-1th acquisition; i Bc represents the occurrence rate of burr sign in the i-th acquisition; i Indicates the boundary clarity value collected for the i-th time; i means starting from the first data, and n means ending at the n-th data.
[0020] Preferably, the internal echo data analysis node is used to establish an internal echo data calculation model, import the internal echo data transmitted by the breast cancer ultrasound image data acquisition module into the internal echo data calculation model; obtain the internal echo characteristic value; the internal echo data calculation model is specifically expressed as:
[0021] ,
[0022] in, represents the internal echo characteristic value calculated for the i-th time; Gd i Indicates the grayscale standard deviation of the i-th acquisition; Cn i Ad represents the number of calcification foci collected for the i-th time; i Indicates the attenuation degree of the rear echo collected for the i-th time; i means starting from the first data, and n means ending at the n-th data.
[0023] Preferably, the blood flow signal data analysis node is used to establish a blood flow signal data calculation model, and the blood flow signal data transmitted by the breast cancer ultrasound image data acquisition module is imported into the blood flow signal data calculation model; the blood flow signal characteristic value is obtained; the blood flow signal data calculation model is specifically expressed as:
[0024] ,
[0025] in, represents the blood flow signal characteristic value calculated for the i-th time, Sv i represents the peak systolic flow velocity collected at the i-th time; Tm i represents the blood flow classification collected at the i-th time, Tm i-1Indicates the blood flow classification collected at the i-1th time.
[0026] Preferably, the imaging data analysis node is used to establish an imaging data calculation model, import the imaging data transmitted by the breast cancer ultrasound image data acquisition module into the imaging data calculation model; obtain the physical characteristic state value of breast cancer; the imaging data calculation model is specifically expressed as:
[0027] ,
[0028] in Sr represents the physical property state value of breast cancer calculated for the i-th time; i Sw represents the spatial resolution of the i-th acquisition; i represents the shear wave velocity collected at the ith time, Sw i-1 represents the shear wave velocity collected at the i-1th time; Em i Indicates the maximum elastic modulus collected for the i-th time, Em i-1 Indicates the maximum value of the elastic modulus collected for the i-1th time; i means starting from the first data, and n means ending at the nth data.
[0029] Preferably, the contrast data analysis node is used to establish a contrast data calculation model, and the contrast data transmitted by the breast cancer ultrasound image data acquisition module is imported into the contrast data calculation model to obtain the tumor hemodynamic characteristic value; the contrast data calculation model is specifically expressed as:
[0030] ,
[0031] in represents the tumor hemodynamic characteristic value calculated for the i-th time; Pt i represents the peak time of the ith collection, Pt i-1 represents the peak time of the i-1th acquisition; Ac i represents the area under the curve of the i-th acquisition; Er i Represents the enhancement intensity ratio of the i-th acquisition.
[0032] Preferably, the breast cancer ultrasound image data evaluation unit is used to construct a breast cancer ultrasound image data calculation model, and the tumor morphology characteristic value, internal echo characteristic value, blood flow signal characteristic value, breast cancer physical property state value and tumor hemodynamic characteristic value transmitted by the breast cancer ultrasound image data analysis module are introduced into the breast cancer ultrasound image data calculation model to obtain a breast cancer comprehensive evaluation index value; the breast cancer ultrasound image data calculation model is specifically expressed as:
[0033] ,
[0034] Among them, I irepresents the comprehensive evaluation index value of breast cancer calculated for the i-th time; represents the tumor morphological characteristic value calculated for the i-th time; represents the internal echo characteristic value calculated for the i-th time; represents the characteristic value of the blood flow signal collected for the i-th time; represents the physical property state value of breast cancer collected for the i-th time; represents the tumor hemodynamic characteristic value calculated for the i-th time; 0.25 represents the weight value of the tumor morphology characteristic value and the internal echo characteristic value; 0.2 represents the weight value of the blood flow signal characteristic value; 0.15 represents the physical property state value of breast cancer and the tumor hemodynamic characteristic value; It represents the standard deviation of tumor morphological characteristic values, internal echo characteristic values, blood flow signal characteristic values, breast cancer physical property state values, and tumor hemodynamic characteristic values; i represents starting from the first data, and n represents ending at the nth data.
[0035] Technical effects and advantages of the present invention:
[0036] The present invention divides the data into different batches according to specific imaging conditions and lesion characteristics through a data batch division module, thereby improving the pertinence and practicality of image analysis and providing an orderly and structured basis for subsequent data processing; the breast cancer ultrasound image data acquisition module acquires data from multiple dimensions, including tumor morphology boundary feature data, internal echo data, blood flow signal data, imaging data and angiography data. This multi-dimensional data acquisition method reduces the one-sidedness of the data and lays a solid foundation for subsequent analysis work;
[0037] The present invention uses a breast cancer ultrasound image analysis module to deeply analyze the collected data, extract key characteristic values, and obtain tumor morphology characteristic values, internal echo characteristic values, blood flow signal characteristic values, breast cancer physical property state values, and tumor hemodynamic characteristic values; this data intuitively reveals the potential problems in breast cancer ultrasound images and provides doctors with a powerful basis for diagnosis;
[0038] The present invention integrates and evaluates the results obtained by the breast cancer ultrasound image analysis module through a comprehensive data evaluation module, further improving the accuracy and practicality of the analysis results, comprehensively considering the correlation and mutual influence between multiple characteristic values, and providing a more comprehensive and objective evaluation report; the analysis results are displayed to doctors in real time in an intuitive manner through a real-time feedback module, and an early warning signal is issued when an abnormality or potential risk is found, so that doctors can timely understand the patient's breast cancer condition and carry out effective management and adjustment of treatment plans. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 It is a schematic diagram of the overall structure of the present invention. DETAILED DESCRIPTION
[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0041] As attached Figure 1 The breast cancer ultrasound image analysis system shown is based on artificial intelligence and includes a data batch division module, a breast cancer ultrasound image data acquisition module, a breast cancer ultrasound image data analysis module, a comprehensive data evaluation module, and a real-time feedback module.
[0042] The data batch division module is used to determine the data to be collected as target data, divide the target data into different batches in an equal time division manner, and mark them as 1, 2, ..., n in sequence;
[0043] What needs to be specifically explained in this embodiment is that the equal time division method is used to divide each target data according to its characteristics through the background information management system, and each target data is divided into different collection times according to its real-time requirements, and a system database is established to store it.
[0044] The breast cancer ultrasound image data acquisition module includes a diagnostic data acquisition unit and an ultrasound data acquisition unit. The diagnostic data acquisition unit is used to acquire tumor morphological boundary feature data, internal echo data, and blood flow signal data; the ultrasound data acquisition unit is used to acquire imaging data and angiography data, and is used to acquire target data in real time and transmit the acquired data to the breast cancer ultrasound image data analysis module.
[0045] In this embodiment, it should be specifically explained that the tumor morphology boundary feature data includes the aspect ratio Ar, the burr sign occurrence rate Mo and the boundary clarity value Bc; the internal echo data includes the grayscale standard deviation Gd, the number of calcification foci Cn and the rear echo attenuation degree Ad; the blood flow signal data includes the systolic peak flow velocity Sv and the blood flow grade Tm.
[0046] In this embodiment, it should be specifically explained that the imaging data includes the spatial resolution Sr, the shear wave velocity Sw and the maximum elastic modulus Em; the angiography data includes the peak time Pt, the area under the curve Ac and the enhancement intensity ratio Er.
[0047] The aspect ratio reflects whether the shape of the tumor is regular. Usually, the aspect ratio of malignant tumors is larger. The acquisition method is: on the ultrasound image, use the measurement software provided by the ultrasound instrument to automatically select the maximum diameter of the tumor parallel to the skin and the maximum diameter of the maximum longitudinal diameter perpendicular to the skin, and then calculate the ratio of the two; the burr sign occurrence rate is an important ultrasound manifestation of malignant tumors, and the acquisition method is: combine the pattern recognition algorithm to observe whether there are pseudopodia or burr signs on the edge of the tumor, count the number of ultrasound images with such features, and calculate their proportion of the total number of images to obtain the occurrence rate; the boundary clarity value is an important prompt of malignant tumors, and the acquisition method is: automatically segment the boundary of the tumor through the edge detection algorithm, analyze the pixel grayscale changes in the boundary area, calculate the width of the edge blur area and convert it into millimeter units through the pixel size of the ultrasound image. The grayscale standard deviation reflects the complexity of the internal structure of the tumor, and its acquisition method is: select the tumor area on the ultrasound image, and use image processing software to calculate the standard deviation of the grayscale of all pixels in the area; the number of calcification foci is a point-like or small-piece area with higher brightness in the ultrasound image, which is of great significance for the diagnosis of malignant tumors, and its acquisition method is: automatically identify and count the number of calcification foci in the tumor through threshold segmentation; the degree of rear echo attenuation is a common ultrasound manifestation of malignant tumors, and its acquisition method is: on the ultrasound image, measure the echo intensity of the rear of the tumor and the normal tissue area respectively, and then calculate the difference between the two, and convert it into decibel units. The peak systolic velocity reflects the blood flow velocity in the tumor blood vessels, which is of great significance for evaluating the blood supply of the tumor, and its acquisition method is: using color Doppler ultrasound technology, select a main blood vessel or an area rich in blood flow signals in the tumor, measure and record the peak systolic velocity; the blood flow grading is one of the important indicators for evaluating the angiogenesis of tumors, and its acquisition method is: according to the Adler grading standard, the ultrasound doctor observes the richness of the blood flow signal in the tumor and performs grading.The spatial resolution directly affects the visibility of image details and is crucial for the detection of tiny lesions. The acquisition method is: using a high-frequency three-dimensional ultrasonic transducer to obtain a high-resolution three-dimensional image of breast tissue by finely adjusting the focus and scanning path of the ultrasonic beam; the shear wave velocity is an important basis for distinguishing benign from malignant lesions, and the acquisition method is: applying ultrasound elastic imaging technology to apply slight vibration to breast tissue through a probe to measure the shear wave velocity generated by the tissue; the maximum elastic modulus reflects the hardness of the tissue and helps to distinguish benign from malignant lesions. The acquisition method is: using ultrasound elastic imaging technology to measure the maximum elastic modulus value in the lesion area; the peak time reflects the blood perfusion rate and vascular permeability in the lesion area, which is of great significance for evaluating tumor angiogenesis, and the acquisition method is: during ultrasound contrast imaging, recording the time when the contrast agent reaches the maximum concentration in the lesion area; the area under the curve helps to evaluate the invasiveness and prognosis of the tumor, and the acquisition method is: calculating the area under the ultrasound contrast imaging time intensity curve; the enhanced intensity ratio reflects the ability of the lesion area to absorb the contrast agent, and the acquisition method is: comparing the gray value changes in the lesion area before and after ultrasound contrast imaging to calculate the enhanced intensity ratio.
[0048] The breast cancer ultrasound image data analysis module includes a diagnostic data analysis unit and an ultrasound data analysis unit, which are used to analyze the data transmitted by the breast cancer ultrasound image data acquisition module and transmit the analysis results to the comprehensive data evaluation module; the diagnostic data analysis unit includes a tumor morphology boundary feature data analysis node, an internal echo data analysis node and a blood flow signal data analysis node; the ultrasound data analysis unit includes an imaging data analysis node and an angiography data analysis node.
[0049] In this embodiment, it should be specifically explained that: the tumor morphology boundary feature data analysis node is used to establish a tumor morphology boundary feature data calculation model, and the tumor morphology boundary feature data transmitted by the breast cancer ultrasound image data acquisition module is imported into the tumor morphology boundary feature data calculation model to obtain a tumor morphology feature value; the tumor morphology boundary feature data calculation model is specifically expressed as:
[0050] ,
[0051] in, Ar represents the tumor morphological characteristic value calculated for the i-th time; i Ar represents the aspect ratio of the i-th acquisition; i-1 Mo represents the aspect ratio of the i-1th acquisition; i Bc represents the occurrence rate of burr sign in the i-th acquisition; i Indicates the boundary clarity value collected for the i-th time; i means starting from the first data, and n means ending at the n-th data.
[0052] In this embodiment, it is specifically required to be explained that: the internal echo data analysis node is used to establish an internal echo data calculation model, and the internal echo data transmitted by the breast cancer ultrasound image data acquisition module is imported into the internal echo data calculation model; the internal echo characteristic value is obtained; the internal echo data calculation model is specifically expressed as:
[0053] ,
[0054] in, represents the internal echo characteristic value calculated for the i-th time; Gd i Indicates the grayscale standard deviation of the i-th acquisition; Cn i Ad represents the number of calcification foci collected for the i-th time; i Indicates the attenuation degree of the rear echo collected for the i-th time; i means starting from the first data, and n means ending at the n-th data.
[0055] In this embodiment, it is specifically required to explain that: the blood flow signal data analysis node is used to establish a blood flow signal data calculation model, and the blood flow signal data transmitted by the breast cancer ultrasound image data acquisition module is imported into the blood flow signal data calculation model; the blood flow signal characteristic value is obtained; the blood flow signal data calculation model is specifically expressed as:
[0056] ,
[0057] in, represents the blood flow signal characteristic value calculated for the i-th time, Sv i represents the peak systolic flow velocity collected at the i-th time; Tm i represents the blood flow classification collected at the i-th time, Tm i-1 Indicates the blood flow classification collected at the i-1th time.
[0058] In this embodiment, it is specifically required to be explained that: the imaging data analysis node is used to establish an imaging data calculation model, and the imaging data transmitted by the breast cancer ultrasound image data acquisition module is imported into the imaging data calculation model; the physical characteristic state value of breast cancer is obtained; the imaging data calculation model is specifically expressed as:
[0059] ,
[0060] in Sr represents the physical property state value of breast cancer calculated for the i-th time; i Sw represents the spatial resolution of the i-th acquisition; i represents the shear wave velocity collected at the ith time, Sw i-1 represents the shear wave velocity collected at the i-1th time; Em i Indicates the maximum elastic modulus collected for the i-th time, Emi-1 Indicates the maximum value of the elastic modulus collected for the i-1th time; i means starting from the first data, and n means ending at the nth data.
[0061] In this embodiment, it is specifically required to be explained that: the contrast data analysis node is used to establish a contrast data calculation model, and the contrast data transmitted by the breast cancer ultrasound image data acquisition module is imported into the contrast data calculation model to obtain the tumor hemodynamic characteristic value; the contrast data calculation model is specifically expressed as:
[0062] ,
[0063] in represents the tumor hemodynamic characteristic value calculated for the i-th time; Pt i represents the peak time of the ith collection, Pt i-1 represents the peak time of the i-1th acquisition; Ac i represents the area under the curve of the i-th acquisition; Er i Represents the enhancement intensity ratio of the i-th acquisition.
[0064] The comprehensive data evaluation module includes a breast cancer ultrasound image data evaluation unit, which comprehensively analyzes the data transmitted by the breast cancer ultrasound image data analysis module and transmits the analysis results to the real-time feedback module.
[0065] In this embodiment, it should be specifically explained that the breast cancer ultrasound image data evaluation unit is used to establish a breast cancer ultrasound image data calculation model, and the tumor morphology characteristic value, internal echo characteristic value, blood flow signal characteristic value, breast cancer physical property state value and tumor hemodynamic characteristic value transmitted by the breast cancer ultrasound image data analysis module are imported into the breast cancer ultrasound image data calculation model to obtain a breast cancer comprehensive evaluation index value; the breast cancer ultrasound image data calculation model is specifically expressed as:
[0066] ,
[0067] Among them, I i represents the comprehensive evaluation index value of breast cancer calculated for the i-th time; represents the tumor morphological characteristic value calculated for the i-th time; represents the internal echo characteristic value calculated for the i-th time; represents the characteristic value of the blood flow signal collected for the i-th time; represents the physical property state value of breast cancer collected for the i-th time; represents the tumor hemodynamic characteristic value calculated for the i-th time; 0.25 represents the weight value of the tumor morphology characteristic value and the internal echo characteristic value; 0.2 represents the weight value of the blood flow signal characteristic value; 0.15 represents the physical property state value of breast cancer and the tumor hemodynamic characteristic value; It represents the standard deviation of tumor morphological characteristic values, internal echo characteristic values, blood flow signal characteristic values, breast cancer physical property state values, and tumor hemodynamic characteristic values; i represents starting from the first data, and n represents ending at the nth data.
[0068] The real-time feedback module is used to establish a preset value of the breast cancer comprehensive evaluation index, judge the breast cancer comprehensive evaluation index value according to the preset value of the breast cancer comprehensive evaluation index, generate a diagnosis report according to the judgment result and save the data in a database.
[0069] In this embodiment, it should be specifically noted that the preset value of the comprehensive evaluation index of breast cancer is marked as 1 def , when I i def When , it means that the breast cancer comprehensive evaluation index value is less than the preset value of the breast cancer comprehensive evaluation index, indicating that there is no breast cancer, and a normal signal is issued; when I i def When , it means that the breast cancer comprehensive assessment index value is greater than or equal to the preset value of the breast cancer comprehensive assessment index, indicating that breast cancer has been contracted, and an early warning signal is issued to relevant personnel, and the breast cancer comprehensive assessment index value is recorded and uploaded to the database.
[0070] The present invention divides the data into different batches according to specific imaging conditions and lesion characteristics through a data batch division module, thereby improving the pertinence and practicality of image analysis and providing an orderly and structured basis for subsequent data processing; the breast cancer ultrasound image data acquisition module acquires data from multiple dimensions, including tumor morphology boundary feature data, internal echo data, blood flow signal data, imaging data and angiography data. This multi-dimensional data acquisition method reduces the one-sidedness of the data and lays a solid foundation for subsequent analysis work; the breast cancer ultrasound image analysis module conducts in-depth analysis on the acquired data, extracts key feature values, and obtains tumor morphology feature values, internal echo feature values, blood flow signal data, imaging data and angiography data. The data directly reveal the potential problems in breast cancer ultrasound images and provide doctors with a strong basis for diagnosis. The comprehensive data evaluation module integrates and evaluates the results of the breast cancer ultrasound image analysis module, further improving the accuracy and practicality of the analysis results. The module comprehensively considers the correlation and mutual influence between multiple characteristic values to provide a more comprehensive and objective evaluation report. The real-time feedback module displays the analysis results to doctors in an intuitive way in real time, and issues early warning signals when abnormalities or potential risks are found, so that doctors can timely understand the patient's breast cancer status and conduct effective management and adjustment of treatment plans.
[0071] Secondly: In the drawings of the embodiments disclosed in the present invention, only the structures related to the embodiments disclosed in the present invention are involved, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other;
[0072] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A breast cancer ultrasound image analysis system based on artificial intelligence, characterized in that: include: Data batch division module: used to determine the data to be collected as target data, divide the target data into different batches according to equal time division, and mark them as 1, 2, ..., n in sequence; Breast cancer ultrasound image data acquisition module: including a diagnostic data acquisition unit and an ultrasound data acquisition unit, wherein the diagnostic data acquisition unit is used to acquire tumor morphological boundary feature data, internal echo data, and blood flow signal data; the ultrasound data acquisition unit is used to acquire imaging data and angiography data, and is used to acquire target data in real time, and transmit the acquired data to the breast cancer ultrasound image data analysis module; The tumor morphological boundary feature data include the aspect ratio Ar, the burr sign occurrence rate Mo and the boundary clarity value Bc; the internal echo data include the grayscale standard deviation Gd, the number of calcification foci Cn and the degree of posterior echo attenuation Ad; the blood flow signal data include the systolic peak flow velocity Sv and the blood flow grade Tm; the imaging data include the spatial resolution Sr, the shear wave velocity Sw and the maximum value of the elastic modulus Em; the angiography data include the peak time Pt, the area under the curve Ac and the enhancement intensity ratio Er; Breast cancer ultrasound image data analysis module: includes a diagnostic data analysis unit and an ultrasound data analysis unit, which is used to analyze the data transmitted by the breast cancer ultrasound image data acquisition module and transmit the analysis results to the comprehensive data evaluation module; the diagnostic data analysis unit includes a tumor morphology boundary feature data analysis node, an internal echo data analysis node, and a blood flow signal data analysis node; the ultrasound data analysis unit includes an imaging data analysis node and an angiography data analysis node; The tumor morphology boundary feature data analysis node is used to establish a tumor morphology boundary feature data calculation model, and import the tumor morphology boundary feature data transmitted by the breast cancer ultrasound image data acquisition module into the tumor morphology boundary feature data calculation model to obtain a tumor morphology feature value; the tumor morphology boundary feature data calculation model is specifically expressed as: , in, Ar represents the tumor morphological characteristic value calculated for the i-th time; i Ar represents the aspect ratio of the i-th acquisition; i-1 Mo represents the aspect ratio of the i-1th acquisition; i Bc represents the occurrence rate of burr sign in the i-th acquisition; i Indicates the boundary definition value collected for the i-th time; i means starting from the first data, and n means ending at the n-th data; The internal echo data analysis node is used to establish an internal echo data calculation model, import the internal echo data transmitted by the breast cancer ultrasound image data acquisition module into the internal echo data calculation model, and obtain the internal echo characteristic value; the internal echo data calculation model is specifically expressed as: , in, represents the internal echo characteristic value calculated for the i-th time; Gd i Indicates the grayscale standard deviation of the i-th acquisition; Cn i Ad represents the number of calcification foci collected for the i-th time; i Indicates the attenuation degree of the rear echo collected for the i-th time; i means starting from the first data, and n means ending at the n-th data; The blood flow signal data analysis node is used to establish a blood flow signal data calculation model, import the blood flow signal data transmitted by the breast cancer ultrasound image data acquisition module into the blood flow signal data calculation model; obtain the blood flow signal characteristic value; the blood flow signal data calculation model is specifically expressed as: , in, represents the blood flow signal characteristic value calculated for the i-th time, Sv i represents the peak systolic flow velocity collected at the i-th time; Tm i represents the blood flow classification collected at the i-th time, Tm i-1 represents the blood flow classification collected at the i-1th time; The imaging data analysis node is used to establish an imaging data calculation model, import the imaging data transmitted by the breast cancer ultrasound image data acquisition module into the imaging data calculation model; obtain the physical characteristic state value of breast cancer; the imaging data calculation model is specifically expressed as: , in Sr represents the physical property state value of breast cancer calculated for the i-th time; i Sw represents the spatial resolution of the i-th acquisition; i represents the shear wave velocity collected at the ith time, Sw i-1 represents the shear wave velocity collected at the i-1th time; Em i Indicates the maximum elastic modulus collected for the i-th time, Em i-1 Indicates the maximum value of the elastic modulus collected at the i-1th time; i indicates starting from the first data, and n indicates ending at the nth data; The contrast data analysis node is used to establish a contrast data calculation model, import the contrast data transmitted by the breast cancer ultrasound image data acquisition module into the contrast data calculation model, and obtain the tumor hemodynamic characteristic value; the contrast data calculation model is specifically expressed as: , in represents the tumor hemodynamic characteristic value calculated for the i-th time; Pt i represents the peak time of the ith collection, Pt i-1 represents the peak time of the i-1th acquisition; Ac i represents the area under the curve of the i-th acquisition; Er i represents the enhancement intensity ratio of the i-th acquisition; Comprehensive data evaluation module: including a breast cancer ultrasound image data evaluation unit, which comprehensively analyzes the data transmitted by the breast cancer ultrasound image data analysis module and transmits the analysis results to the real-time feedback module; The breast cancer ultrasound image data evaluation unit is used to establish a breast cancer ultrasound image data calculation model, and import the tumor morphology characteristic value, internal echo characteristic value, blood flow signal characteristic value, breast cancer physical property state value and tumor hemodynamic characteristic value transmitted by the breast cancer ultrasound image data analysis module into the breast cancer ultrasound image data calculation model to obtain a breast cancer comprehensive evaluation index value; the breast cancer ultrasound image data calculation model is specifically expressed as: , Among them, I i represents the comprehensive evaluation index value of breast cancer calculated for the i-th time; represents the tumor morphological characteristic value calculated for the i-th time; represents the internal echo characteristic value calculated for the i-th time; represents the characteristic value of the blood flow signal collected for the i-th time; represents the physical property state value of breast cancer collected for the i-th time; represents the tumor hemodynamic characteristic value calculated for the i-th time; 0.25 represents the weight value of the tumor morphology characteristic value and the internal echo characteristic value; 0.2 represents the weight value of the blood flow signal characteristic value; 0.15 represents the physical property state value of breast cancer and the tumor hemodynamic characteristic value; Represents the standard deviation of tumor morphological characteristic values, internal echo characteristic values, blood flow signal characteristic values, breast cancer physical property state values, and tumor hemodynamic characteristic values; i means starting from the first data, and n means ending at the nth data; Real-time feedback module: used to establish preset values of the comprehensive evaluation index of breast cancer, judge the comprehensive evaluation index value of breast cancer according to the preset values of the comprehensive evaluation index of breast cancer, generate a diagnosis report according to the judgment result and save the data in the database.
Citation Information
Patent Citations
Breast cancer prediction model based on multi-parameter ultrasound and construction method thereof
CN114999640A
