Medical data processing system and method

Abstract

The present invention relates to the field of medical ultrasonic image processing technology, and specifically to a medical data processing system and method. The present invention can quickly acquire high-quality original ultrasonic images through an efficient data acquisition unit, laying a solid foundation for subsequent processing. The image analysis unit uses frequency analysis technology to accurately identify image features, thereby improving the accuracy and reliability of diagnosis. The filtering and noise reduction unit effectively eliminates interference information in the frequency image, enhancing the clarity and readability of the image. Finally, the imaging unit reconstructs the processed image into a high-quality target image, providing an intuitive and detailed medical reference for subsequent ultrasonic image processing. The present invention improves the efficiency and quality of medical image processing, can effectively reduce the risk of misdiagnosis caused by human factors, provides strong support for clinical diagnosis and treatment, and helps promote the intelligent development of the medical industry.

Description

Technical Field

[0001] The present invention relates to the technical field of medical ultrasonic image processing, and in particular to a medical data processing system and method. Background Art

[0002] Medical ultrasound imaging is a practical medical imaging technology that enables non-invasive visualization of soft tissues through the propagation and reflection of sound waves.

[0003] Medical ultrasound image segmentation aims to automatically identify and extract targets of interest in ultrasound images, such as body tissues, organs, or lesion areas, thereby improving diagnostic efficiency and accuracy. Accurate segmentation can improve the reliability of quantitative analysis of ultrasound images, reduce the workload of doctors, and provide support for subsequent diagnosis and surgery.

[0004] However, the original ultrasound images often contain a large amount of stray noise information, which hinders the identification of key areas and subsequent quantitative analysis. The ultrasound images need to be processed before use. Traditional image segmentation methods have many problems, such as loose structure, poor information fusion between models, insufficient feature integration of deep learning segmentation models, limited expression of multi-scale semantic information, and high computational cost, which lead to low image segmentation accuracy and low segmentation efficiency, and therefore need to be solved urgently. Summary of the invention

[0005] In order to avoid and overcome the technical problems existing in the prior art, the present invention provides a medical data processing system and method. The present invention can effectively reduce the noise of ultrasonic images so as to improve the accuracy of subsequent image processing.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A medical data processing system comprises the following components:

[0008] Data acquisition unit: used to acquire original ultrasound images;

[0009] Image analysis unit: used for frequency analysis of ultrasound images;

[0010] Filtering and noise reduction unit: used for filtering and noise reduction processing of ultrasonic images;

[0011] Imaging unit: used to reconstruct the ultrasonic image after filtering and noise reduction into a target image.

[0012] A medical data processing method, which uses the above-mentioned medical data processing system, includes the following processing steps:

[0013] S1, acquiring an ultrasound image through a data acquisition unit, and performing Fourier transform on the ultrasound image to obtain a frequency domain image;

[0014] S2, using an image analysis unit to count the distribution of frequency types in the frequency domain image, and selecting a filtering method from low-pass filtering, high-pass filtering and band-pass filtering as a reference filtering according to the distribution;

[0015] S3, in the filtering and denoising unit, respectively and simultaneously use low-pass filtering, high-pass filtering and band-pass filtering to perform filtering and denoising processing on the frequency domain image, and perform inverse Fourier transform on the frequency domain image after filtering and denoising processing to obtain a filtered image;

[0016] S4. In the imaging unit, the pixel value of each pixel point in the filtered image is determined by threshold processing based on the reference filter. The filtered image after the pixel value is determined is the target image.

[0017] As a further solution of the present invention: the reference filter selection process is as follows:

[0018] S21, counting the frequency types in the frequency domain image in the image analysis unit and storing them in a frequency set, and selecting the maximum frequency value and the minimum frequency value from the frequency set;

[0019] S22, calculating the absolute value of the frequency difference between each frequency in the frequency set and the minimum frequency value and summing them, and recording the summation result as the minimum difference sum; at the same time, calculating the absolute value of the frequency difference between each frequency in the frequency set and the maximum frequency value and summing them, and recording the summation result as the maximum difference sum;

[0020] S23, comparing the minimum difference sum and the maximum difference sum, and selecting a corresponding filtering method as a reference filtering in combination with a reference filtering determination equation.

[0021] As a further solution of the present invention: the reference filtering determination equation is expressed as follows:

[0022] ;

[0023] ;

[0024] ;

[0025] Where, RPF represents reference filtering; LPF represents low-pass filtering; HPF represents high-pass filtering; BPF represents band-pass filtering; MAX represents the maximum difference sum; MIN represents the minimum difference sum; N represents the total number of frequency categories in the frequency concentration; f min Indicates the minimum frequency in the frequency concentration; f max Indicates the maximum frequency value in the frequency set.

[0026] As a further solution of the present invention: the cut-off frequency of the low-pass filter is calculated as follows:

[0027] ;

[0028] In the formula, f LPF Indicates the cutoff frequency of the low-pass filter;

[0029] The cutoff frequency of the high-pass filter is calculated as follows:

[0030] ;

[0031] In the formula, f HPF Represents the cutoff frequency of the high-pass filter;

[0032] The endpoint values ​​of the bandpass filter bandwidth are calculated as follows;

[0033] ;

[0034] In the formula, f down Indicates the left endpoint value of the bandpass filter bandwidth; f up Indicates the right endpoint value of the bandpass filter bandwidth; Represents the average frequency in the frequency set.

[0035] As a further solution of the present invention: the process of acquiring the target image is as follows:

[0036] S41, constructing a filter weight calculation equation in the imaging unit, and calculating the filter weights of low-pass filtering, high-pass filtering and band-pass filtering respectively according to the filter weight calculation equation;

[0037] S42, respectively obtaining the original pixel values ​​of each pixel point in the three filtered images, performing weighted processing on the original pixel values ​​by using a pixel weighting equation to obtain a weighted pixel value, and using the weighted pixel value as the pixel value of the current pixel point;

[0038] S43, obtaining the filtered image processed in step S42, where the filtered image is the desired target image.

[0039] As a further solution of the present invention: the filter weight calculation equation is expressed as follows:

[0040] ;

[0041] In the formula, w LPF represents the filter weight of low-pass filtering; w HPF represents the filter weight of high-pass filtering; w BPF Represents the filter weight of the bandpass filter.

[0042] As a further solution of the present invention: the pixel weighting equation is expressed as follows:

[0043] ;

[0044] In the formula, p i represents the pixel value of the i-th pixel in the target image, that is, the weighted pixel value; p i,LPF represents the original pixel value of the i-th pixel in the filtered image obtained by low-pass filtering; p i,HPF represents the original pixel value of the i-th pixel in the filtered image obtained by high-pass filtering; p i,BPF Represents the original pixel value of the i-th pixel in the filtered image obtained by bandpass filtering.

[0045] As a further solution of the present invention: the process of acquiring the frequency domain image is as follows:

[0046] S11, selecting an ultrasound image of a corresponding human body part from historical ultrasound images retained in a historical detection process of the ultrasound device through a data acquisition unit, and storing the image in an ultrasound data set;

[0047] S12, converting each ultrasound image in the ultrasound data set from the spatial domain to the frequency domain using a two-dimensional fast Fourier transform;

[0048] S13, performing visualization processing on the ultrasound image converted to the frequency domain in step S12, so as to obtain a corresponding frequency domain image.

[0049] As a further solution of the present invention: the process of obtaining the filtered image is as follows:

[0050] S31, in the filtering and denoising unit, respectively and simultaneously use low-pass filtering, high-pass filtering and band-pass filtering with determined parameters to perform filtering and denoising processing on the frequency domain image;

[0051] S32, performing a two-dimensional inverse fast Fourier transform on the frequency domain image after the filtering and noise reduction processing in step S31 to obtain a corresponding filtered image.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. The system of the present invention integrates multiple key functions such as data acquisition, image analysis, filtering and noise reduction, and imaging, and realizes the comprehensive optimization of medical image processing. Through the efficient data acquisition unit, the system can quickly obtain high-quality original ultrasound images, laying a solid foundation for subsequent processing. The image analysis unit uses advanced frequency analysis technology to accurately identify image features and improve the accuracy and reliability of diagnosis. The filtering and noise reduction unit effectively eliminates interference information in the image and enhances the clarity and readability of the image. Finally, the imaging unit reconstructs the processed image into a high-quality target image, providing an intuitive and detailed medical reference for subsequent ultrasound image processing. The system not only improves the efficiency and quality of medical image processing, but also reduces the risk of misdiagnosis caused by human factors, provides strong support for clinical diagnosis and treatment, and helps promote the intelligent development of the medical industry.

[0054] 2. The method of the present invention achieves in-depth optimization of ultrasonic images by cleverly combining data acquisition, frequency domain analysis, multiple filtering techniques and threshold processing. First, the use of Fourier transform converts the image from the time domain to the frequency domain, which is convenient for subsequent frequency distribution statistics and filter selection. Secondly, the reference filtering method is intelligently selected according to the frequency distribution to ensure the pertinence and effectiveness of the filtering process. At the same time, low-pass filtering, high-pass filtering and band-pass filtering are applied in parallel, supplemented by inverse Fourier transform, which greatly enriches the flexibility and accuracy of image processing. Finally, threshold processing based on reference filtering further improves the clarity and detail expression of the target image, providing more reliable and detailed image support for medical diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a schematic diagram of the structure of the medical data processing system in the present invention.

[0056] Figure 2 This is a flow chart of the medical data processing method in the present invention. DETAILED DESCRIPTION

[0057] 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.

[0058] See also Figure 1 The medical data processing system of the present invention is described in detail as follows, and aims to achieve accurate acquisition, in-depth analysis, effective noise reduction and high-quality imaging of ultrasonic images through highly integrated and intelligent technical means, so as to provide strong support for medical diagnosis.

[0059] 1. Data acquisition unit.

[0060] 1.1. Ultrasonic probe assembly.

[0061] Function: As the front-end sensing device of the system, it is responsible for emitting ultrasonic waves and receiving reflected signals from human tissues.

[0062] Features: It uses high-precision, multi-frequency optional ultrasound probe to ensure that the resolution and depth of image acquisition meet clinical needs.

[0063] 1.2. Data acquisition circuit.

[0064] Function: Convert the analog signal received by the ultrasound probe into a digital signal and perform preliminary preprocessing.

[0065] Features: Integrated high-speed analog-to-digital converter, supports high-resolution, low-noise data acquisition, and ensures the accuracy and integrity of image data.

[0066] The ultrasonic probe is used to perform ultrasonic detection on characteristic parts of the human body, and the detected data is transmitted to the data acquisition circuit. The data acquisition circuit converts the ultrasonic signal into a digital signal and stores it in a database.

[0067] 2. Image analysis unit.

[0068] 2.1. Signal amplification and filtering components.

[0069] Function: Amplify and preliminarily filter the collected digital signals to remove high-frequency noise and interference.

[0070] Features: Adopt programmable gain amplifier and analog filter to achieve dynamic range adjustment and noise suppression of digital signals.

[0071] 2.2. Image format conversion component.

[0072] Function: Convert the pre-processed digital signal into a standard image format, that is, form an ultrasound image, which is convenient for subsequent processing and analysis.

[0073] Features: Supports multiple image formats (such as BMP, JPEG, DICOM, etc.) to ensure compatibility with existing medical systems.

[0074] 2.3. Frequency analysis component.

[0075] Function: Perform Fourier transform on the preprocessed ultrasound image and analyze the frequency components in the ultrasound image.

[0076] Features: Use two-dimensional fast Fourier transform to achieve efficient and accurate frequency analysis.

[0077] The digital signal in the database is amplified and preliminarily filtered by the signal amplification and filtering component, then converted into an ultrasound image by the image format conversion component, and finally the frequency type of the ultrasound image is analyzed by the frequency analysis component.

[0078] 3. Filtering and noise reduction unit.

[0079] 3.1. Filter.

[0080] Function: According to the analysis results of the frequency types in the ultrasound image, select the appropriate filtering method (such as low-pass filtering, high-pass filtering, band-pass filtering, etc.).

[0081] Features: Supports multiple filtering methods and can flexibly configure corresponding filters according to actual needs.

[0082] 4. Imaging unit.

[0083] 4.1. Image reconstruction component.

[0084] Function: Perform inverse Fourier transform on the filtered image data to reconstruct a clear and accurate medical image, i.e., the target image.

[0085] Features: Adopt high-quality image reconstruction algorithm to ensure image resolution and detail expression.

[0086] 4.2. Image enhancement and adjustment components.

[0087] Function: Further enhance the reconstructed medical images, such as contrast adjustment, sharpening, color correction, etc.

[0088] Features: Provides a wealth of image enhancement tools that can be personalized according to actual needs.

[0089] 5. System control and interaction module.

[0090] 5.1. Display.

[0091] Function: Provides an intuitive and easy-to-use user interface for users to easily perform system settings, image preview and result viewing.

[0092] Features: Graphical interface design is adopted, supporting touch and mouse operation to improve user experience.

[0093] 5.2. System controller.

[0094] Function: Responsible for the overall control and scheduling of the system to ensure the coordinated work between various units.

[0095] Features: Use embedded systems or high-performance computers as the control core to achieve real-time and efficient system control.

[0096] In summary, the medical data processing system achieves accurate acquisition, in-depth analysis, effective noise reduction and high-quality imaging of ultrasound images through highly integrated and intelligent technical means, providing strong support for medical diagnosis. At the same time, the system is also highly flexible and scalable, and can be customized and optimized according to actual needs.

[0097] See also Figure 2 The medical data processing method of the present invention comprises the following steps:

[0098] 1. Acquire an ultrasound image through a data acquisition unit and perform Fourier transform on it to obtain a frequency domain image.

[0099] S11. Select 10 ultrasound images of the same liver part from the historical ultrasound images retained in the historical detection process of the ultrasound equipment through the data acquisition unit, and store them in the ultrasound data set.

[0100] S12. Use two-dimensional fast Fourier transform to convert each ultrasound image in the ultrasound data set from the spatial domain to the frequency domain.

[0101] S13, performing visualization processing on the ultrasound image converted to the frequency domain in step S12, so as to obtain a clear frequency domain image.

[0102] Second, the distribution of frequency types in the frequency domain image is statistically analyzed by the image analysis unit, and a filtering method is selected from low-pass filtering, high-pass filtering and band-pass filtering as a reference filtering according to the distribution.

[0103] S21. In the image analysis unit, each ultrasound image of the liver is processed separately. Taking an ultrasound image as an example, the frequency types in the frequency domain image of the ultrasound image are counted and stored in a frequency set. There are 10 different frequency values ​​in the frequency set, namely 1Hz, 5Hz, 55Hz, 65Hz, 66Hz, 70Hz, 73Hz, 76Hz, 78Hz and 80Hz, among which the maximum frequency is 80Hz, the minimum frequency is 1Hz, and the average frequency is 56.9Hz.

[0104] S22, calculate the absolute value of the frequency difference between each frequency in the frequency set and the minimum frequency value and sum them, and record the sum result as the minimum difference sum, which is 559 Hz. At the same time, calculate the absolute value of the frequency difference between each frequency in the frequency set and the maximum frequency value and sum them, and record the sum result as the maximum difference sum, which is 231 Hz.

[0105] S23, comparing the minimum difference sum and the maximum difference sum: 559 Hz is greater than 231 Hz, and selecting the corresponding high-pass filter as the reference filter in combination with the reference filter determination equation.

[0106] The cutoff frequency of the low-pass filter is calculated as follows:

[0107] ;

[0108] The cutoff frequency of the high-pass filter is calculated as follows:

[0109] ;

[0110] The endpoint values ​​of the bandpass filter bandwidth are calculated as follows;

[0111] ;

[0112] ;

[0113] 3. In the filtering and noise reduction unit, low-pass filtering, high-pass filtering and band-pass filtering are used simultaneously to filter the frequency domain image, and the frequency domain image after filtering is subjected to inverse Fourier transform to obtain a filtered image.

[0114] S31 , in a filtering and noise reduction unit, respectively and simultaneously use low-pass filtering, high-pass filtering and band-pass filtering with determined parameters to perform filtering processing on the frequency domain image.

[0115] S32, performing a two-dimensional inverse fast Fourier transform on the frequency domain image after the filtering process in step S31 to obtain a corresponding filtered image.

[0116] Fourth, in the imaging unit, the pixel value of each pixel point in the filtered image is determined by threshold processing based on the reference filter, and the filtered image after the pixel value is determined is the target image.

[0117] S41, constructing a filter weight calculation equation in the imaging unit, and respectively calculating the filter weights of low-pass filtering, high-pass filtering and band-pass filtering according to the filter weight calculation formula.

[0118] ;

[0119] ;

[0120] S42, respectively obtain the original pixel value of each pixel in the three filtered images. Take the pixel value of one pixel for illustration, and analyze the remaining pixels in the same way. The original pixel value of the low-pass filter is 20; the original pixel value of the high-pass filter is 21; and the original pixel value of the band-pass filter is 18.

[0121] The original pixel value is weighted by the pixel weighting equation to obtain the weighted pixel value. The weighted pixel value of the pixel point is:

[0122] ;

[0123] The weighted pixel value of the pixel is rounded off to an integer, and is obtained as 20, and the weighted pixel value is used as the pixel value of the current pixel.

[0124] S43, obtaining the filtered image processed in step S42, where the filtered image is the desired target image.

[0125] Through the above embodiments, the various steps of the medical data processing method are verified in detail, including obtaining frequency domain images, selecting reference filtering methods, filtering processing, and determining target images. This process shows how to extract clearer and more useful information from the original ultrasound image so that doctors can make more accurate diagnoses.

[0126] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A medical data processing method, characterized in that: The processing steps include: S1, acquiring an ultrasound image through a data acquisition unit, and performing Fourier transform on the ultrasound image to obtain a frequency domain image; S2, using an image analysis unit to count the distribution of frequency types in the frequency domain image, and selecting a filtering method from low-pass filtering, high-pass filtering and band-pass filtering as a reference filtering according to the distribution; The reference filter selection process is as follows: S21, counting the frequency types in the frequency domain image in the image analysis unit and storing them in a frequency set, and selecting the maximum frequency value and the minimum frequency value from the frequency set; S22, calculating the absolute value of the frequency difference between each frequency in the frequency set and the minimum frequency value and summing them, and recording the summation result as the minimum difference sum; at the same time, calculating the absolute value of the frequency difference between each frequency in the frequency set and the maximum frequency value and summing them, and recording the summation result as the maximum difference sum; S23, comparing the minimum difference sum and the maximum difference sum, and selecting the corresponding filtering method as the reference filtering in combination with the reference filtering determination equation; the reference filtering determination equation is expressed as follows: ； ； ； In the formula, RPF represents reference filtering; LPF Represents low-pass filtering; HPF Represents high-pass filtering; BPF represents bandpass filtering; MAX represents the maximum difference sum; MIN represents the minimum sum of differences; N Indicates the total number of frequency categories in the frequency concentration; f min Indicates the minimum frequency value in the frequency concentration; f max Indicates the maximum frequency in the frequency concentration; S3, in the filtering and noise reduction unit, respectively and simultaneously use low-pass filtering, high-pass filtering and band-pass filtering to filter the frequency domain image, and perform inverse Fourier transform on the filtered frequency domain image to obtain a filtered image; S4, determining the pixel value of each pixel point in the filtered image by threshold processing based on the reference filter in the imaging unit, and the filtered image after the pixel value is determined is the target image; The process of acquiring the target image is as follows: S41, constructing a filter weight calculation equation in the imaging unit, and calculating the filter weights of low-pass filtering, high-pass filtering and band-pass filtering respectively according to the filter weight calculation formula; The filter weight calculation equation is expressed as follows: ； In the formula, w LPF Represents the filter weight of low-pass filtering; w HPF Represents the filter weight of high-pass filtering; w BPF represents the filter weight of bandpass filtering; S42, respectively obtaining the original pixel values ​​of each pixel point in the three filtered images, performing weighted processing on the original pixel values ​​by using a pixel weighting equation to obtain a weighted pixel value, and using the weighted pixel value as the pixel value of the current pixel point; The pixel weighting equation is expressed as follows: ； In the formula, p i Indicates the target image i The pixel value of a pixel point, that is, the weighted pixel value; p i,LPF Indicates the number of filtered images obtained by low-pass filtering. i The original pixel value of each pixel; p i,HPF Indicates the number of filtered images obtained by high-pass filtering. i The original pixel value of each pixel; p i,BPF Indicates the first i The original pixel value of each pixel; S43, obtaining the filtered image processed in step S42, where the filtered image is the desired target image.

2. A medical data processing method according to claim 1, characterized in that: The cutoff frequency of the low-pass filter is calculated as follows: ； In the formula, f LPF Indicates the cutoff frequency of the low-pass filter; The cutoff frequency of the high-pass filter is calculated as follows: ； In the formula, f HPF Represents the cutoff frequency of the high-pass filter; The endpoint values ​​of the bandpass filter bandwidth are calculated as follows; ； In the formula, f down Indicates the left endpoint value of the bandpass filter bandwidth; f up Indicates the right endpoint value of the bandpass filter bandwidth; Represents the average frequency in the frequency set.

3. A medical data processing method according to claim 2, characterized in that: The process of acquiring the frequency domain image is as follows: S11, selecting a plurality of ultrasound images of the same human body part from the historical ultrasound images retained in the historical detection process of the ultrasound equipment through the data acquisition unit, and storing them in the ultrasound data set; S12, converting each ultrasound image in the ultrasound data set from the spatial domain to the frequency domain using a two-dimensional fast Fourier transform; S13, performing visualization processing on the ultrasound image converted to the frequency domain in step S12, so as to obtain a corresponding frequency domain image.

4. A medical data processing method according to claim 3, characterized in that: The process of obtaining the filtered image is as follows: S31, in the filtering and noise reduction unit, respectively and simultaneously use low-pass filtering, high-pass filtering and band-pass filtering with determined parameters to filter the frequency domain image; S32, performing a two-dimensional inverse fast Fourier transform on the frequency domain image after the filtering process in step S31 to obtain a corresponding filtered image.

5. A medical data processing system, the medical data processing system applying a medical data processing method as claimed in any one of claims 1 to 4, characterized in that: It includes the following components: Data acquisition unit: used to acquire original ultrasound images; Image analysis unit: used for frequency analysis of ultrasound images; Filtering and noise reduction unit: used for filtering and noise reduction processing of ultrasonic images; Imaging unit: used to reconstruct the filtered ultrasound image into a target image.

Images