Portable ultrasonic cardiovascular assessment system and method

Abstract

The invention discloses a portable ultrasonic cardiovascular assessment system and method, and the system comprises (a) an ultrasonic probe which is internally provided with an FPGA module and is used for carrying out the ROI positioning and compressed LBP feature extraction in real time; (b) a host processor configured to receive the compressed LBP features and perform dynamic puzzle recombination; (c) a rhythm adaptive switching module for detecting an RR interval variation coefficient gt; automatically starting nonlinear EF calculation at 15%; according to the portable ultrasonic equipment, the adaptability and precision of EF calculation are improved through the technologies of myocardial texture feature extraction, nonlinear cardiac cycle recombination, multi-scale time-frequency analysis and the like, the portable ultrasonic equipment is particularly suitable for patients with arrhythmia, in addition, the portable ultrasonic equipment can efficiently and accurately carry out EF measurement, and the requirements of clinical rapid diagnosis and remote medical treatment are met.

Description

Technical Field

[0001] The present invention belongs to the field of medical image processing and cardiovascular function evaluation, and specifically relates to a portable ultrasonic cardiovascular evaluation system and method. Background Art

[0002] Cardiovascular diseases (CVDs) are one of the leading causes of death and disability globally. Among them, the left ventricular ejection fraction (EF) is an important indicator for evaluating cardiac function. In clinical practice, EF is usually evaluated by echocardiography to determine the heart's pumping ability and assist in the diagnosis and management of heart failure, coronary heart disease, and other cardiovascular diseases.

[0003] Currently, the calculation of EF in echocardiography mainly relies on the following methods: Biplane Simpson's method: This method calculates based on the change in left ventricular volume at the end of cardiac systole and diastole and is applicable to two-dimensional ultrasound images. However, this method has the following problems: It depends on the boundary tracking of the left ventricle and is easily affected by the operator's experience, resulting in poor repeatability. In patients with arrhythmia, the morphological changes in consecutive cardiac cycles are relatively large, making it difficult to accurately determine the end points of systole and diastole, and the calculation error is relatively large.

[0004] M-mode ultrasound EF measurement method: This method calculates the EF value by measuring the change in the anteroposterior diameter of the left ventricle and is applicable to patients with relatively regular left ventricular function. However, in the presence of myocardial motion abnormalities (such as segmental motion abnormalities, arrhythmia, etc.), the accuracy of this method drops significantly.

[0005] EF calculation based on tissue Doppler imaging (TDI): In recent years, TDI has been used for EF calculation, and cardiac function is deduced based on parameters such as myocardial velocity and strain rate. However, this method is limited by the imaging quality and noise interference of ultrasonic equipment, and has high requirements for image processing and computing capabilities, and is not applicable to portable ultrasonic devices. Summary of the Invention

[0006] The purpose of the present invention is to provide a portable ultrasonic cardiovascular evaluation system and method. The present invention improves the adaptability and accuracy of EF calculation through technologies such as myocardial texture feature extraction, non-linear cardiac cycle recombination, and multi-scale time-frequency analysis, especially applicable to patients with arrhythmia. In addition, through optimized designs such as real-time calculation by the FPGA module and feature-compressed data transmission, the portable ultrasonic device can efficiently and accurately measure EF, meeting the needs of clinical rapid diagnosis and telemedicine.

[0007] The technical solution adopted by the present invention is as follows: A method for calculating the cardiovascular ejection fraction based on dynamic recombination of myocardial texture, comprising the following steps: (a) Real-time collect a sequence of cardiac ultrasound images and automatically locate the region of interest (ROI) of the left ventricular free wall; (b) Extract the myocardial texture features of the ROI region to generate a dynamic texture waveform; (c) Detect the systolic start and diastolic end events in the texture waveform through multi-scale time-frequency analysis; (d) When arrhythmia is detected, non-linearly recombine the image fragments of different cardiac cycles according to the events; (e) Calculate the non-uniform ejection fraction (EF) based on the recombined dynamic jigsaw puzzle.

[0008] Among them, the myocardial texture feature in step (b) is the compressed local binary pattern (LBP), which is generated in the following way: (i) Calculate the standard LBP code for each block of the ROI region; (ii) Perform Gray code conversion on the LBP code; (iii) Use parallel Hamming weight to calculate the compressed feature dimension.

[0009] Among them, the multi-scale time-frequency analysis in step (c) uses the stepped wavelet transform.

[0010] Among them, the non-linear recombination in step (d) includes: (i) Construct a similarity matrix of cardiac cycle fragments, where the similarity calculation is: ; Among them: represents the similarity score between the and cardiac cycle fragments, is the Pearson correlation coefficient function, is the compressed local binary pattern feature vector of the th fragment, is the timestamp of the corresponding fragment (unit: second); (ii) Use the greedy algorithm to connect the fragments with similarity exceeding the threshold.

[0011] Among them, the calculation formula for the non-uniform EF in step (e) is: ; Among them: A max,i is the maximum area of the myocardial texture in the i-th cardiac cycle (unit: pixel ; A min,i is the minimum area of the myocardial texture in the i-th cardiac cycle (unit: pixel ; is the median function, which is used to extract representative values in multiple cycles; k is the volume - area correction factor, and its value range is 0.85 - 0.95, which is calibrated through the clinical dataset.

[0012] A portable ultrasonic cardiovascular assessment system for implementing the method, comprising: (a) An ultrasonic probe with an in - built FPGA module for real - time execution of ROI positioning and compressed LBP feature extraction; (b) A host processor configured to receive the compressed LBP features and perform dynamic puzzle recombination; (c) A rhythm - adaptive switching module that automatically enables non - linear EF calculation when the coefficient of variation of the RR interval > 15% is detected.

[0013] Among them, the FPGA module uses a look - up table (LUT) to solidify the weights of the transfer learning model, where: (i) The ROI positioning model is a lightweight version of MobileNetV2; (ii) The fully - connected layer is replaced by bilinear interpolation approximate calculation.

[0014] Among them, the data transmission between the ultrasonic probe and the host uses a feature compression protocol, including: (a) Compress the 256×256 pixels of the original ultrasonic image into a 112 - bit feature vector; (b) Use differential coding to reduce redundant data in consecutive frames.

[0015] A computer - readable storage medium stores a computer program, which implements the steps of the method when the program is executed by a processor.

[0016] An ultrasonic image processing device includes a memory, a processor, and a computer program stored on the memory. The processor implements the steps of the method when executing the program.

[0017] In summary, due to the adoption of the above - mentioned technical solutions, the beneficial effects of the present invention are: 1. Existing ultrasonic EF (ejection fraction) measurements are usually based on left ventricular volume calculation, while the present invention realizes a more refined assessment through dynamic recombination of myocardial texture. Compressed local binary pattern (LBP) is used to extract myocardial texture features, which can capture local changes in myocardial contraction more accurately compared with traditional boundary - tracking methods.

[0018] 2. Existing methods have large errors in patients with arrhythmia, while the present invention detects systolic and diastolic events in the cardiac cycle through multi - scale time - frequency analysis and can adapt to different rhythm changes. By using a non - linear recombination method, different cardiac cycle segments are recombined by constructing a similarity matrix and a greedy algorithm, improving the accuracy of EF calculation in the case of arrhythmia.

[0019] 3. The stepped wavelet transform is used for multi-scale time-frequency analysis, which can capture the dynamic changes of myocardial texture more effectively compared with the traditional FFT (Fast Fourier Transform). Based on the volume-area correction factor of the patient's body type, the accuracy of personalized EF calculation can be improved compared with the fixed model.

[0020] 4. Existing portable ultrasound devices usually have limited computing power. In the present invention, the combination of the FPGA module and the host processor is used to improve the computing efficiency: the ultrasound probe is built-in with an FPGA module for real-time ROI positioning and compressed LBP feature extraction, reducing the computing burden of the host. A rhythm adaptive switching module is adopted to automatically enable non-linear EF calculation when the RR interval variation > 15%, improving the applicability in case of arrhythmia.

[0021] 5. Existing ultrasound devices have a large amount of data, which affects transmission and storage. In the present invention, a look-up table (LUT) is used to solidify the weights of the transfer learning model to optimize the computing burden. The feature compression protocol compresses the ultrasound image from 256×256 pixels to a 112-bit feature vector, and differential coding is used to reduce the data redundancy between frames, improving the real-time performance.

[0022] 6. The present invention is not limited to the method, and an ultrasound image processing device is also proposed, including a memory, a processor, and a computer program, forming a complete software and hardware combination solution. The computer-readable storage medium can be used for subsequent software upgrades and adaptation between different devices, improving the scalability of the solution.

[0023] In summary, compared with the prior art, the present invention has significant advantages in improving the EF calculation accuracy, adapting to arrhythmia, optimizing the computing and storage efficiency, etc. At the same time, combined with the design of the portable ultrasound device, it realizes a more efficient, accurate, and portable cardiovascular assessment. Description of the Drawings

[0024] Figure 1 It is a working schematic diagram of the portable ultrasound cardiovascular assessment system of the present invention; Figure 2 It is a flow schematic diagram of the cardiovascular ejection fraction calculation method of the present invention. Detailed Embodiments

[0025] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0026] See Figure 1 and 2, the present invention relates to a method for calculating cardiovascular ejection fraction based on dynamic reorganization of myocardial texture, comprising the following steps: (a) Real-time collect a sequence of cardiac ultrasound images and automatically locate the region of interest (ROI) of the left ventricular free wall; (b) Extract the myocardial texture features of the ROI region to generate a dynamic texture waveform; (c) Detect the systolic start and diastolic end events in the texture waveform through multi-scale time-frequency analysis; (d) When arrhythmia is detected, non-linearly reorganize the image segments of different cardiac cycles according to the events; (e) Calculate the non-uniform ejection fraction (EF) based on the reorganized dynamic jigsaw puzzle.

[0027] Further, the myocardial texture feature in step (b) is the compressed local binary pattern (LBP), which is generated in the following manner: (i) Calculate the standard LBP coding for each block of the ROI region; (ii) Perform Gray coding conversion on the LBP coding; (iii) Calculate the compressed feature dimension using parallel Hamming weight.

[0028] Further, the multi-scale time-frequency analysis in step (c) adopts the stepped wavelet transform.

[0029] Further, the non-linear reorganization in step (d) includes: (i) Construct a similarity matrix for cardiac cycle segments, where the similarity calculation is: ; Where: represents the similarity score between the and cardiac cycle segments, is the Pearson correlation coefficient function, is the compressed local binary pattern feature vector of the segment, is the time stamp of the corresponding segment (unit: second); (ii) Use the greedy algorithm to connect the segments with similarity exceeding the threshold.

[0030] Further, the formula for calculating the non-uniform EF in step (e) is: ; Where: A max,i is the maximum area of myocardial texture in the i-th cardiac cycle (unit: pixel ; A min,i is the minimum area of myocardial texture in the i-th cardiac cycle (unit: pixel ; is the median function, which is used to extract representative values in multiple cycles; k is the volume-area correction factor, and its value range is 0.85 to 0.95. Calibrated by the clinical data set, the volume-area correction factor (k) is an empirical coefficient used in the present invention to estimate the myocardial texture area in the two-dimensional image as the actual volume change of the left ventricle. Since the area change in the image cannot directly reflect the true change of the ventricular volume, it is necessary to multiply the area difference by the correction factor k according to the clinical data statistics. This factor is obtained from the fitting relationship between the large-sample true EF and the texture area, and usually takes values between 0.85 and 0.95. It is an empirical value obtained through data training, not a single fixed value. Under the influence of different patient body types, imaging angles, etc., the k value can be individually adjusted according to the device default model or the clinical calibration database.

[0031] A portable ultrasonic cardiovascular assessment system for implementing the method, comprising: (a) An ultrasonic probe with an internal FPGA module for real-time execution of ROI positioning and compressed LBP feature extraction; (b) A host processor configured to receive the compressed LBP features and perform dynamic puzzle recombination; (c) A rhythm adaptive switching module that automatically enables non-linear EF calculation when the coefficient of variation of the RR interval > 15% is detected.

[0032] Further, the FPGA module uses a look-up table (LUT) to solidify the weights of the transfer learning model, where: (i) The ROI positioning model is a lightweight version of MobileNetV2; (ii) The fully connected layer is replaced by bilinear interpolation approximate calculation.

[0033] Further, the data transmission between the ultrasonic probe and the host uses a feature compression protocol, including: (a) Compressing the 256×256 pixels of the original ultrasonic image into a 112-bit feature vector; (b) Using differential coding to reduce redundant data in consecutive frames.

[0034] A computer-readable storage medium stores a computer program, and when the program is executed by a processor, the steps of the method are implemented.

[0035] An ultrasonic image processing device includes a memory, a processor, and a computer program stored on the memory. When the processor executes the program, the steps of the method are implemented.

[0036] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for calculating cardiovascular ejection fraction based on dynamic reconstruction of myocardial texture, characterized in that: The following steps are involved: (a) Real-time acquisition of cardiac ultrasound image sequences and automatic location of the region of interest (ROI) of the left ventricular free wall; (b) extracting myocardial texture features of the ROI region and generating a dynamic texture waveform; (c) detecting the systolic onset and diastolic end events in the texture waveform by multi-scale time-frequency analysis; (d) when arrhythmia is detected, nonlinearly recombining image segments of different cardiac cycles according to the event; (e) Calculation of heterogeneous ejection fraction (EF) based on the reorganized dynamic puzzle.

2. The method according to claim 1, characterized in that: The myocardial texture feature in step (b) is a compressed local binary pattern (LBP), which is generated by: (i) Calculate the standard LBP coding for the ROI area blocks; (ii) performing Gray coding conversion on the LBP code; (iii) Compress feature dimensions using parallel Hamming weight computation.

3. The method according to claim 1, characterized in that The multi-scale time-frequency analysis in step (c) uses stepped wavelet transform.

4. The method according to claim 1, characterized in that: The nonlinear reorganization in step (d) comprises: (i) Construct a similarity matrix of cardiac cycle segments, where the similarity is calculated as: ； in: Indicates and The similarity score of each cardiac cycle segment is is the Pearson correlation coefficient function, For the The compressed local binary pattern feature vector of the segments, is the timestamp of the corresponding fragment (unit: seconds); (ii) A greedy algorithm is used to connect the segments whose similarity exceeds a threshold.

5. The method according to claim 1, characterized in that The calculation formula of non-uniform EF in step (e) is: ； Among them: A max,i is the maximum area of ​​myocardial texture in the ith cardiac cycle (unit: pixel ; A min,i is the minimum area of ​​myocardial texture in the ith cardiac cycle (unit: pixel ; is the median function, which is used to extract representative values ​​under multiple cycles; k is the volume-area correction factor, which ranges from 0.85 to 0.95 and is calibrated by the clinical data set.

6. A portable ultrasonic cardiovascular assessment system, used to implement the method according to any one of claims 1 to 5, characterized in that: include: (a) Ultrasound probe with built-in FPGA module for real-time ROI localization and compressed LBP feature extraction; (b) a host processor configured to receive the compressed LBP features and perform dynamic jigsaw puzzle reconstruction; (c) The rhythm adaptation switching module automatically enables nonlinear EF calculation when the RR interval variation coefficient is detected to be >15%.

7. The system according to claim 6, characterized in that The FPGA module uses a lookup table (LUT) to solidify the transfer learning model weights, where: (i) The ROI positioning model is a lightweight version of MobileNetV2; (ii) The fully connected layer is replaced by bilinear interpolation approximation.

8. The system according to claim 6, characterized in that The data transmission between the ultrasound probe and the host adopts a characteristic compression protocol, including: (a) The 256×256 pixels of the original ultrasound image are compressed into a 112-bit feature vector; (b) Use differential coding to reduce redundant data of consecutive frames.

9. A computer-readable storage medium storing a computer program, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

10. An ultrasonic image processing device, comprising a memory, a processor and a computer program stored in the memory, characterized in that: When the processor executes the program, the steps of the method according to any one of claims 1 to 5 are implemented.

Images