A method, apparatus and electronic device for calculating blood flow images

Abstract

The present invention relates to the technical field of image processing, and in particular to a blood flow image calculation method, device and electronic device. The method includes: obtaining a set of scanned images, where the set of scanned images includes at least two OCT images obtained by scanning a target position; for each of the OCT images, performing Fourier transform on the OCT image to obtain a complex amplitude image; based on at least two preset convolution kernels, performing convolution calculation on each convolution kernel and each complex amplitude image. In this application, convolution calculation is performed on the complex amplitude image after Fourier transform through a convolution kernel, and then calculation is performed according to multiple complex images corresponding to each convolution kernel, taking into account the amplitude and phase relationship between adjacent pixels to improve the accuracy of the calculation result. Then, a blood flow image is calculated according to all the first target results to fully optimize the signal-to-noise ratio of the blood flow image, thereby improving the accuracy of the blood flow image.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular, to a blood flow calculation method, device, electronic device, and storage medium for optical coherence tomography. Background Art

[0002] Optical coherence tomography angiography (OCT Angiography) is an important non-invasive technology for diagnosing fundus diseases that has emerged in recent years.

[0003] Currently, most blood flow technology solutions are based on continuously acquiring multiple B-scan images at a single scanning position and analyzing the differences between these B-scan images to obtain blood flow information. Most of the existing analysis methods are one-to-one mappings where a single pixel in the output corresponds to the same position in the input image, without considering the possible relationships between the amplitudes and phases of adjacent pixels, resulting in room for improvement in the accuracy of the analysis results. Summary of the Invention

[0004] In view of this, an object of the present invention is to provide a blood flow calculation method, device, and electronic device.

[0005] In a first aspect, an embodiment of the present invention provides a blood flow image calculation method, the method comprising:

[0006] Obtaining a set of scanned images, the set of scanned images including at least two OCT images obtained by scanning a target position;

[0007] For each of the OCT images, performing a Fourier transform on the OCT image to obtain a complex amplitude image;

[0008] Based on at least two preset convolution kernels, performing a convolution calculation on each convolution kernel and each complex amplitude image to obtain a set of sub-images corresponding to each convolution kernel;

[0009] Determining the first image set according to all the sets of sub-images;

[0010] Calculating the images in the first image set to obtain a blood flow image.

[0011] Combined with the first aspect, the step of calculating the images in the first image set to obtain a blood flow image includes:

[0012] Based on a preset pairing rule, pairing the images in the set of sub-images corresponding to each convolution kernel to obtain image pairs;

[0013] For each of the image pairs, performing a decorrelation calculation on the image pair to obtain a first intermediate result corresponding to the image pair;

[0014] Determine a first target result corresponding to each of the convolution kernels according to all the first intermediate results;

[0015] Calculate a blood flow image according to all the first target results.

[0016] Combined with the first aspect, for each of the image pairs, the step of performing decorrelation calculation on the image pair includes:

[0017] Calculate according to the following formula:

[0018] D (a，a+1，i) = ||CK (a，i) |-|CK (a+1，i) || / (|CK (a，i) |+|CK (a+1，i) |+Sigma);

[0019] where D (a，a+1，i) is the first intermediate result obtained by performing decorrelation calculation on the image pair formed by the a-th complex amplitude image and the (a + 1)-th complex amplitude image after convolution calculation with the i-th convolution kernel respectively; CK (a，i) is the complex image obtained by convolving the a-th complex amplitude image with the i-th convolution kernel, CK (a+1，i) is the complex image obtained by convolving the (a + 1)-th complex amplitude image with the i-th convolution kernel, and Sigma is zero or a positive number.

[0020] Combined with the first aspect, for each of the image pairs, the step of performing decorrelation calculation on the image pair includes:

[0021] Calculate according to the following formula:

[0022] D (a，a+1，i) = (|CK (a，i) |-|CK (a+1，i) |) 2 / (|CK (a，i) | 2 +|CK (a+1，i) | 2 +Sigma).

[0023] Combined with the first aspect, after the step of pairing the images in the sub-image set corresponding to each of the convolution kernels based on a preset pairing rule to obtain image pairs, it further includes:

[0024] For each of the image pairs, determine whether the absolute values of the first complex image and the second complex image in the image pair are both less than a first set threshold;

[0025] If so, determine that the first intermediate result corresponding to the image pair is a preset value.

[0026] In combination with the first aspect, the steps of calculating the blood flow image according to all the first target results include:

[0027] Obtain the weights corresponding to each convolutional kernel;

[0028] Sum the products of the weights corresponding to each convolutional kernel and the first target results corresponding to the convolutional kernels to obtain the blood flow image.

[0029] In combination with the first aspect, after the step of obtaining the scanned image set, the following steps are further included:

[0030] Input the scanned image set into a preset model to output the target blood flow image.

[0031] In combination with the first aspect, the training process of the model is as follows:

[0032] Obtain training samples, where the training samples include a first number of images;

[0033] Calculate the first blood flow image according to the first number of the images by a preset algorithm;

[0034] Use the first blood flow image as the target result, and initialize the convolutional kernel and weights;

[0035] Select a second number of images from the first number of images as a second image set; wherein, the second number is less than the first number;

[0036] According to the initialized convolutional kernel and the weights, calculate the images in the second image set by using the blood flow image calculation method to obtain a second blood flow image;

[0037] Calculate the residual between the second blood flow image and the target result;

[0038] Accumulate the residuals to obtain the loss function of the model;

[0039] Optimize the model according to the loss function until a preset condition is met to obtain the optimized model, where the optimized model includes the optimized convolutional kernel and the weights corresponding to the optimized convolutional kernel.

[0040] In a second aspect, the present application provides a blood flow image calculation device, and the device includes:

[0041] An image acquisition module, configured to acquire a scanned image set, where the scanned image set includes at least two OCT images obtained by scanning a target position;

[0042] An image transformation module, configured to perform Fourier transform on each OCT image to obtain a complex amplitude image;

[0043] A convolution calculation module, configured to perform convolution calculation on each convolution kernel and each complex amplitude image based on at least two preset convolution kernels to obtain a subset of images corresponding to each convolution kernel;

[0044] A determination module, configured to determine the first set of images according to all the subsets of images;

[0045] A blood flow image calculation module, configured to calculate the images in the first set of images to obtain a blood flow image.

[0046] In a third aspect, the present application provides an electronic device, including: a memory and a processor, where the memory stores a computer program, and the processor is configured to execute the computer program to implement the method as described above.

[0047] The embodiments of the present invention bring the following beneficial effects: The present invention provides a method, apparatus and electronic device for calculating a blood flow image. The method includes: obtaining a set of scanned images, where the set of scanned images includes at least two OCT images obtained by scanning a target position; for each OCT image, performing Fourier transform on the OCT image to obtain a complex amplitude image; based on at least two preset convolution kernels, performing convolution calculation on each convolution kernel and each complex amplitude image to obtain a subset of images corresponding to each convolution kernel; determining the first set of images according to all the subsets of images; calculating the images in the first set of images to obtain a blood flow image.

[0048] In the present application, convolution calculation is performed on the complex amplitude image after Fourier transform through a convolution kernel, and then calculation is performed according to multiple complex images corresponding to each convolution kernel, taking into account the amplitude and phase relationship between adjacent pixels to improve the accuracy of the calculation result. Then, a blood flow image is calculated according to all the first target results to fully optimize the signal-to-noise ratio of the blood flow image, thereby improving the accuracy of the blood flow image.

[0049] Other features and advantages of the present invention will be described in the following specification, and part of them will become obvious from the specification or be understood by implementing the present invention. The objectives and other advantages of the present invention are achieved and obtained by the structures specifically pointed out in the specification, claims and drawings.

[0050] To make the above objectives, features and advantages of the present invention more obvious and understandable, the following specific embodiments are given and described in detail in conjunction with the accompanying drawings. Description of the Drawings

[0051] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those skilled in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0052] Figure 1 Flowchart of the blood flow image calculation method provided by an embodiment of the present invention;

[0053] Figure 2 Schematic structural diagram of the blood flow image calculation device provided by an embodiment of the present invention;

[0054] Figure 3 Schematic structural diagram of the electronic device provided by an embodiment of the present invention;

[0055] Figure 4 The first blood flow image obtained by calculating the collected B-scan image according to a preset algorithm in the related art;

[0056] Figure 5 The second blood flow image obtained by performing convolution calculation on the collected B-scan image by the method provided by an embodiment of the present invention.

[0057] Reference numerals:

[0058] 10 - Image acquisition module, 20 - Image transformation module, 30 - Convolution calculation module, 40 - Determination module, 50 - Blood flow image calculation module;

[0059] 41 - Processor, 42 - Bus, 43 - Communication interface, 44 - Memory. Specific embodiments

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention with reference to the drawings. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0061] To facilitate the understanding of this embodiment, the following will first briefly introduce the technical terms designed in this application.

[0062] Optical Coherence Tomography (OCT) technology has been widely used in the diagnosis of fundus diseases and is of great significance for the detection and treatment of ophthalmic diseases. By using the coherence of light to scan and image the fundus, each scan in the depth direction at the same lateral position is called an A-scan, and a combination of adjacent consecutive scans is called a B-scan. A B-scan image is also the OCT cross-sectional image commonly seen. The OCT image is an image obtained by scanning the target position with an OCT device, and this OCT image can be an A-scan image or a B-scan image. Compared with traditional methods, this method not only does not require a blood flow contrast agent, but also can clearly image large and small blood vessels and capillaries in multiple physiological structure layers within the retina. It can be used to diagnose various fundus diseases such as macular degeneration, neovascularization, diabetic retinopathy, vascular occlusion, central serous chorioretinopathy, and so on.

[0063] After introducing the technical terms related to this application, next, a brief introduction to the application scenario and design concept of the embodiments of this application will be given.

[0064] Existing analysis methods are mostly one-to-one mappings where one output pixel corresponds to the same position of a single pixel in the input image, without considering the possible relationships between the amplitudes and phases of statistically adjacent pixels.

[0065] Based on this, the embodiments of this application provide a blood flow image calculation method, device, and electronic device. After obtaining a set of scanned images, convolution calculation is performed on each B-scan image in the set of scanned images through a convolution kernel to introduce the mutual relationship between the phases and amplitudes of different pixels around one pixel, and then the final blood flow image is calculated. In this way, the signal-to-noise ratio and accuracy of the blood flow image can be improved to obtain an accurate blood flow calculation result. The blood flow calculation method provided in this embodiment is applied to the processor 41 in the electronic device. The electronic device also includes a memory 44, and a computer program is stored in the memory 44. The processor 41 executes the computer program to implement the method provided in the embodiments of this application.

[0066] Embodiment 1

[0067] This embodiment provides another blood flow image calculation method. As shown in Figure 1 , this method includes:

[0068] S110, the processor obtains a set of scanned images, and the set of scanned images includes at least two OCT images obtained by scanning the target position.

[0069] S120, for each OCT image, the processor performs a Fourier transform on the OCT image to obtain a complex amplitude image.

[0070] S130, the processor performs convolution calculations on each complex amplitude image with at least two preset convolution kernels to obtain a set of sub-images corresponding to each convolution kernel.

[0071] S140, determine the first image set according to all the sets of sub-images.

[0072] S150, the processor calculates the images in the first image set to obtain a blood flow image.

[0073] In this embodiment, step S110 obtains at least two B-scan images by scanning a target position with an OCT device or an OCTA device; as an implementable manner, the scanning position is a set position, the depth direction is defined as the z-axis of the examined eye coordinate system, and the x-axis and y-axis are planned in a plane perpendicular to the z-axis. Along the x-axis in the plane coordinate system as the fast scan direction and the y-axis as the slow scan direction, at least two B-scan images are continuously acquired in the same slow scan direction. In this embodiment, the number M of B-scan images takes a value of 4.

[0074] Combined with the above example, for each OCT image in step S120, performing a Fourier transform on the OCT image can obtain a complex amplitude image corresponding to each image; combined with the above example, M complex amplitude images C i (i = 1,..., M) are obtained.

[0075] The number of preset convolution kernels in step S130 is N, denoted as K i (i = 1,..., N), and these N convolution kernels are respectively subjected to convolution calculations with the M complex amplitude images obtained in step S120 to obtain a set of sub-images corresponding to each convolution kernel. By S130, the amplitude and phase relationships between adjacent pixels are taken into account to improve the accuracy of the calculation results.

[0076] For example, the preset N convolution kernels K j , (j = 1,..., N), N ≥ 2. Here, N = 3 is taken as an example. Then, each of the 4 complex amplitude images C i obtained by the above Fourier transform is subjected to convolution calculation with these N convolution kernels to obtain 4x3 complex number images Here, i = 1, 2, 3, 4; j = 1, 2, 3; is the convolution operation.

[0077] In step S130, each convolution kernel corresponds to a set of sub-images, and this set of sub-images contains M complex number images, with a total of N sets of sub-images.

[0078] Step S140 forms the first image set from the set of N sets of sub-images.

[0079] Step S150: Calculate the images in the first image set obtained after convolution calculation to obtain the final blood flow image.

[0080] Compared with the prior art, in this embodiment, by performing convolution on the complex amplitude image after Fourier transform with the convolution kernel, a correlation relationship between the phase and amplitude of one pixel and different surrounding pixels is introduced, and then calculations are performed between different images on the convolution result, thereby improving the signal-to-noise ratio and accuracy of the blood flow image.

[0081] Combined with the first aspect, the step of calculating the images in the first image set in step S150 to obtain the blood flow image specifically includes:

[0082] S151: Based on a preset pairing rule, the processor pairs the images in the sub-image set corresponding to each convolution kernel to obtain image pairs.

[0083] S152: For each image pair, the processor performs decorrelation calculation on the image pair to obtain the first intermediate result corresponding to the image pair.

[0084] S153: The processor determines the first target result corresponding to each convolution kernel according to all the first intermediate results.

[0085] S154: The processor calculates according to all the first target results to obtain the blood flow image.

[0086] In this embodiment, first, decorrelation calculation is performed on each sub-image set through steps S151 - S152 to obtain the first intermediate result corresponding to the sub-image set, then the first result corresponding to each convolution kernel is determined by combining all the first intermediate results through step S153, and finally the blood flow image is calculated by combining all the first target results through step S154.

[0087] Taking i = 3 as an example, the source of the images in the sub-image set corresponding to the convolution kernel K3 is: obtained by performing convolution calculation on M complex amplitude images with the convolution kernel K3. That is, the number of complex images corresponding to the convolution kernel K3 is M. Adjacent two complex images are paired and decorrelation calculation is performed to obtain the first intermediate result between these two complex images. Here, adjacent means adjacent in acquisition time; performing decorrelation on two complex images is essentially performing decorrelation calculation on the corresponding pixel points on the two complex images.

[0088] Considering that the time span between non-adjacent B-scan images is relatively large and eye movement will introduce more noise, it is therefore preferred to perform decorrelation calculation between adjacent B-scan images.

[0089] For the same convolution kernel K1, perform decorrelation calculations between the CK (complex images) given by different B-scan images (i.e., between CK1 and CK2, CK2 and CK3, CK3 and CK4), and obtain the corresponding decorrelation results, which are the first intermediate results.

[0090] Specifically, as an implementable method, S152 performs decorrelation calculations on the image pair to obtain the first intermediate result corresponding to the image pair according to the following formula:

[0091] D (a，a+1，i) = ||CK (a，i) |-|CK (a+1，i) || / (|CK (a，i) |+|CK (a+1，i) |+Sigma);

[0092] As another implementable method, S152 performs decorrelation calculations on the image pair to obtain the first intermediate result corresponding to the image pair according to the following formula:

[0093] D (a，a+1) = (|CK (a，i) |-|CK (a+1，i) |) 2 / (|CK (a，i) | 2 +|CK (a+1，i) | 2 +Sigma).

[0094] Among them, D (a，a+1，i) is the first intermediate result obtained by performing decorrelation calculations on the image pair composed of the a-th complex amplitude image and the (a + 1)-th complex amplitude image after convolution calculations with the i-th convolution kernel respectively; i is the i-th convolution kernel, CK (a，i) is the complex image obtained by convolving the a-th complex amplitude image with the i-th convolution kernel, CK (a+1，i) is the complex image obtained by convolving the (a + 1)-th complex amplitude image with the i-th convolution kernel, and Sigma is zero or a positive number, selected according to the noise of the system. When the OCT image signal intensity is very low, the CK value is very low and D approaches 0.

[0095] As another implementable method, after the step of S151 obtaining the image pair, it further includes:

[0096] S1510, for each image pair, the processor determines whether the absolute values of the first complex image and the second complex image in the image pair are both less than the first set threshold.

[0097] If so, execute step S1511.

[0098] S1511, the processor determines that the first intermediate value corresponding to the image pair is a preset value.

[0099] In this embodiment, the preset value is 0.

[0100] The above three methods can all be implemented; the third method can also be implemented in combination with the first or second method; it is not limited here.

[0101] After that, in step S153, according to all the first intermediate results, the first target result corresponding to each convolution kernel is determined. The first target result corresponding to each convolution kernel can be determined by taking the average result or the median of all the first intermediate results corresponding to each convolution kernel, and it is not limited here.

[0102] Combined with the above example, decorrelation calculation is performed between C1 and C2 to obtain D (1，2) , and D is also calculated in the same way (2，3) , D (3，4) .

[0103] After that, calculate the first target result D = (D (1，2) + D (2，3) + D (3，4) ) ÷ 3.

[0104] Another example is listed: when there are 2 B-scan images in the acquired scan image set and the number of convolution kernels is 1, the number of corresponding image pairs is 1. At this time, the calculated first intermediate result is the first target result.

[0105] Combined with the first aspect, the steps of calculating the blood flow image according to all the first target results in step S154 specifically include:

[0106] S1541, the processor obtains the weight corresponding to each convolution kernel.

[0107] S1542, the processor sums the products of the weight corresponding to each convolution kernel and the first target result corresponding to the convolution kernel to obtain the blood flow image.

[0108] In step S1541, each convolution kernel corresponds to a weight. In this embodiment, the weights W i and the convolution kernels K i are preset according to experience.

[0109] Obtain the preset weights W i (i = 1,..., N) corresponding to each convolution kernel; the final blood flow image is obtained by summing the products of the weight corresponding to each convolution kernel and the first target result corresponding to the convolution kernel. In this way, the signal-to-noise ratio of the blood flow image can be fully optimized, thereby improving the accuracy of the blood flow image.

[0110] In combination with the first aspect, step S154 calculates based on all the first target results to obtain the blood flow image, and specifically may further include:

[0111] Calculate the average result of all the first target results, and the average result is used as the final blood flow image.

[0112] Embodiment 2

[0113] In the blood flow image calculation method provided in this embodiment, the weight W i and the convolution kernel K i are obtained according to preset learning.

[0114] Specifically: after step S110, the set of scanned images obtained in step S110 is input into a preset model to output the target blood flow image.

[0115] In this embodiment, the training process of the model is as follows:

[0116] S210, the processor obtains training samples, and the training samples include a first number of images.

[0117] In this embodiment, in this embodiment, OCT data with a relatively large number of acquisition repetitions is collected. For example, each scanning position is repeated 16 times, that is, M = 16 B-scan images are obtained.

[0118] S220, the processor calculates the first number of images according to a preset algorithm to obtain the first blood flow image.

[0119] Among them, the preset algorithm refers to the blood flow image obtained by the blood flow algorithm that has not undergone convolution calculation in the related technology as the first blood flow image angio GT (ground truth, abbreviated as GT).

[0120] In this embodiment, the calculation formula is as follows:

[0121]

[0122] Among them, Angio is the output blood flow image, C i is the i-th complex amplitude image, Sigma is zero or a positive number, and Sigma is determined according to the system noise.

[0123] Due to the multiple repeated acquisitions at each scanning position, the target image usually has a high signal-to-noise ratio. Data with poor signal-to-noise ratio due to eye movement can be manually selected and discarded and not used as training data.

[0124] Furthermore, the result can be subjected to low-pass and denoising processing to further improve the signal-to-noise ratio of the target result.

[0125] S230, the processor takes the first blood flow image as the target result and initializes the convolution kernel and weights.

[0126] The method for initializing the convolution kernel can be as follows: Assuming the number of convolution kernels is n, set the central element of the first convolution kernel to 1 and the rest to 0; set all elements at a distance of 1 from the center of the second convolution kernel to 1 / L, where L is the number of all non-zero elements, and the rest to 0; and so on.

[0127] Initialize the weights where, W i is the weight corresponding to the i-th convolution kernel.

[0128] S240, the processor selects a second number of images from the first number of images as the second image set; wherein, the second number is less than the first number.

[0129] S250, the processor calculates the images in the second image set according to the initialized convolution kernel and weights using the blood flow image calculation method to obtain the second blood flow image.

[0130] The calculation method is: Extract all adjacent 2 B-scan images from the same set of data, and use the convolution-based blood flow image algorithm calculation method proposed in steps S120 - S150 of Embodiment 1 to obtain the second blood flow image.

[0131] S260, the processor calculates the residual between the second blood flow image and the target result.

[0132] S270, the processor accumulates the residuals to obtain the loss function of the model.

[0133] The formula for accumulating the residuals is:

[0134] Loss = ∑|angio - angio _GT |;

[0135] where, angio is the second blood flow image calculated using the convolution-based angio algorithm proposed in steps S110 - S150, and angio _GT is the target image.

[0136] S280, the processor optimizes the model according to the loss function until a preset condition is met to obtain the optimized model, where the optimized model includes the optimized convolution kernel and the weights corresponding to the optimized convolution kernel.

[0137] Using the gradient descent method, minimize this loss function to optimize and obtain N convolution kernels K i and weights W i and store them in the memory.

[0138] Among them, the model for optimization can be a machine learning model, and the optimized convolution kernel can be used as the preset convolution kernel in the next blood flow image calculation method. That is, in practical application of step S130 of the blood flow image calculation method provided by this application, the convolution kernel can be set according to experience or obtained by training the model. If it is obtained by model training, then the processor acquires N convolution kernels K i and weights W i data in the memory, and then executes steps S120 - S150 according to the N convolution kernels K i and weights W i to calculate the blood flow image.

[0139] Based on the above method, Figure 4 exemplarily shows a first blood flow image obtained by calculation according to a preset algorithm (without convolution calculation) at a certain scanning position. Figure 5 The second blood flow image is obtained by performing convolution calculation on the acquired image based on the method provided by the embodiment of this application at the same acquisition position. It can be seen that the method provided by the embodiment of this application is beneficial to improving the signal-to-noise ratio.

[0140] For the i-th convolution kernel, calculate the blood flow image corresponding to this convolution kernel according to the following formula:

[0141] Angio = ∑(D i ×W i )÷∑W i .

[0142] Among them, Angio is the output blood flow image, D i is the first target result corresponding to the i-th convolution kernel, and W i is the weight corresponding to the i-th convolution kernel.

[0143] Combined with the first aspect, before the step of performing Fourier transform on the image in S120, it further includes:

[0144] Performing dispersion correction processing on the image to correct the dispersion of the image.

[0145] Among them, the dispersion correction can include a third-order polynomial correction method. It is a correction method using the phase fitted by a third-order polynomial. The second-order and third-order parameters used in the correction are usually obtained by calibrating the system offline. Since the first-order dispersion correction corresponds to simply translating the Fourier-transformed image, it is generally not corrected. This is a conventional technical means and is not limited here.

[0146] Combined with the first aspect, after the step of performing Fourier analysis on the image in S120 to obtain the complex amplitude image, it further includes:

[0147] S121. For each complex amplitude image, the processor calculates the absolute value of the signal corresponding to the complex amplitude image as the real amplitude image.

[0148] S122. For each real amplitude image, the processor calculates the logarithm of the signal corresponding to the real amplitude image to obtain the first structural image.

[0149] S123. The processor selects one of the first structural images as the reference image and determines the first structural images other than the reference image as the images to be corrected.

[0150] S124. For each image to be corrected, the processor calculates the misalignment amount between the reference image and the image to be corrected.

[0151] S125. The processor corrects the complex amplitude image according to all the misalignment amounts to obtain the corrected complex amplitude image.

[0152] In this embodiment, by taking the absolute value A i corresponding to the signal of the complex amplitude image C i = |C i |, then taking the logarithm to obtain the first structural image; then performing misalignment calibration on the first structural image, and correcting the complex amplitude image according to the calculated misalignment amount.

[0153] This process can be expressed as: S i = 20×log 10 (A i )(i = 1,..., 4); where S i represents the i-th OCT structural image corresponding to the i-th complex amplitude image.

[0154] In actual application, usually, the OCT structural image corresponding to the complex amplitude image obtained by Fourier transform of the first acquired image is selected as the reference image.

[0155] After that, steps S230 - S250 are performed according to the corrected complex amplitude image to further improve the accuracy of data processing.

[0156] In a second aspect, the present application provides a blood flow image calculation device. As shown in Figure 2 , the device includes: an image acquisition module 10, an image transformation module 20, a convolution calculation module 30, a determination module 40, and a blood flow image calculation module 50.

[0157] The image acquisition module 10 is configured to acquire a set of scanned images, and the set of scanned images includes at least two OCT images obtained by scanning a target position.

[0158] The image transformation module 20 is configured to perform a Fourier transform on each OCT image to obtain a complex amplitude image for each OCT image.

[0159] The convolution calculation module 30 is configured to perform convolution calculations on each complex amplitude image with at least two preset convolution kernels to obtain a set of sub-images corresponding to each convolution kernel;

[0160] The determination module 40 is configured to determine a first image set based on all the sets of sub-images.

[0161] The blood flow image calculation module 50 is configured to calculate the images in the first image set to obtain a blood flow image.

[0162] In a third aspect, the present application provides an electronic device. As Figure 3 shown, the electronic device includes a memory 44 and a processor 41. The memory 44 stores a computer program, and the processor 41 executes the computer program to implement the method as described above. As Figure 4 shown, the electronic device further includes a bus 42 and a communication interface 43. Among them, the processor 41, the communication interface 43, and the memory 44 are connected through the bus 42. Among them, the memory 44 may include a high-speed random access memory (RAM, Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory. Through at least one communication interface 43 (which can be wired or wireless), a communication connection is achieved between this system network element and at least one other network element. The Internet, wide area network, local area network, metropolitan area network, etc. can be used. The bus 42 may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc., and may also be an AMBA (Advanced Microcontroller Bus Architecture) bus. Among them, AMBA defines three buses, including an APB (Advanced Peripheral Bus) bus, an AHB (Advanced High-performance Bus) bus, and an AXI (Advanced eXtensible Interface) bus. The bus 42 can be divided into an address bus, a data bus, a control bus, etc. For the sake of easy representation, Figure 4It is represented only by a two-way arrow, but it does not mean that there is only one bus or one type of bus.

[0163] The processor 41 may be an integrated circuit chip with the ability to process signals. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor 41 or the instructions in the form of software. The above-mentioned processor 41 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field programmable gate array (FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 44, and the processor 41 reads the information in the memory 44 and combines its hardware to complete the Figure 1 method shown above.

[0164] In a fourth aspect, the present application provides a computer-readable storage medium. A computer program is stored in the computer-readable storage medium, and the processor 41 executes the computer program to implement the method as described above. For the specific implementation, reference may be made to the method embodiments, which will not be elaborated herein.

[0165] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems and devices described above can refer to the corresponding processes in the foregoing method embodiments, which will not be elaborated herein.

[0166] In addition, in the description of the embodiments of the present invention, unless otherwise clearly defined and limited, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0167] If the above-mentioned functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0168] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0169] Finally, it should be noted that the above embodiments are only specific embodiments of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting them. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A blood flow image calculation method, characterized in that, The method includes: Obtaining a set of scanned images, where the set of scanned images includes at least two OCT images obtained by scanning a target position; For each of the OCT images, performing a Fourier transform on the OCT image to obtain a complex amplitude image; Based on at least two preset convolution kernels, performing convolution calculations on each convolution kernel and each complex amplitude image to obtain a set of sub-images corresponding to each convolution kernel; Determining a first set of images according to all the sets of sub-images; Based on a preset pairing rule, pairing the images in the set of sub-images corresponding to each convolution kernel to obtain image pairs; For each of the image pairs, performing decorrelation calculation on the image pair to obtain a first intermediate result corresponding to the image pair; Determining a first target result corresponding to each convolution kernel according to all the first intermediate results; Calculating a blood flow image according to all the first target results.

2. The method according to claim 1, wherein For each of the image pairs, the step of performing decorrelation calculation on the image pair includes: Calculating according to the following formula: D (a，a+1，i) = ||CK (a，i) |-|CK (a+1，i) || / (|CK (a，i) |+|CK (a+1，i) |+Sigma); Among them, D (a，a+1，i) is the first intermediate result obtained by performing decorrelation calculation on the image pair composed of the a-th complex amplitude image and the (a + 1)-th complex amplitude image after convolution calculation with the i-th convolution kernel respectively; CK (a，i) is the complex image obtained by convolving the a-th complex amplitude image with the i-th convolution kernel, CK (a+1，i) is the complex image obtained by convolving the (a + 1)-th complex amplitude image with the i-th convolution kernel, and Sigma is zero or a positive number.

3. The method according to claim 1, wherein For each of the image pairs, the step of performing decorrelation calculation on the image pair includes: Calculating according to the following formula: D (a，a+1，i) = (|CK (a，i) | - |CK (a+1，i) |) 2 / (|CK (a，i) | 2 + |CK (a+1，i) | 2 + Sigma); Among them, D (a，a+1，i) is the first intermediate result obtained by performing decorrelation calculation on the image pair composed of the a-th complex amplitude image and the (a + 1)-th complex amplitude image after convolution calculation with the i-th convolution kernel respectively; CK (a，i) is the complex image obtained by performing convolution calculation on the a-th complex amplitude image with the i-th convolution kernel, and CK (a+1，i) is the complex image obtained by performing convolution calculation on the (a + 1)-th complex amplitude image with the i-th convolution kernel, and Sigma is zero or a positive number.

4. The method according to claim 1, characterized in that After the step of pairing the images in the set of sub-images corresponding to each convolution kernel based on a preset pairing rule to obtain image pairs, it further includes: For each of the image pairs, determining whether the absolute values of the first complex image and the second complex image in the image pair are both less than a first set threshold; If so, determining that the first intermediate result corresponding to the image pair is a preset value.

5. The method according to claim 1, characterized in that, The step of calculating a blood flow image according to all the first target results includes: Obtaining the weight corresponding to each convolution kernel; Summing the products of the weight corresponding to each convolution kernel and the first target result corresponding to the convolution kernel to obtain the blood flow image.

6. The method according to claim 5, wherein After the step of obtaining the set of scanned images, it further includes: Inputting the set of scanned images into a preset model and outputting the blood flow image.

7. The method according to claim 6, wherein It includes: The training process of the model is as follows: Obtaining training samples, where the training samples include a first number of images; Calculating a first blood flow image according to a preset algorithm for the first number of images; Taking the first blood flow image as the target result and initializing the convolution kernel and the weight; Selecting a second number of images from the first number of images as a second set of images; where the second number is less than the first number; Calculating the images in the second set of images according to the initialized convolution kernel and the weight by using the blood flow image calculation method to obtain a second blood flow image; Calculating the residual between the second blood flow image and the target result; Accumulating the residuals to obtain a loss function; Optimizing the model according to the loss function until a preset condition is met to obtain the optimized model, where the optimized model includes the optimized convolution kernel and the weight corresponding to the optimized convolution kernel.

8. A blood flow image calculation device, characterized in that, The device includes: An image acquisition module for obtaining a set of scanned images, where the set of scanned images includes at least two OCT images obtained by scanning a target position; An image transformation module, configured to perform Fourier transform on each OCT image to obtain a complex amplitude image; A convolution calculation module, configured to perform convolution calculation on each convolution kernel and each complex amplitude image based on at least two preset convolution kernels to obtain a subset of images corresponding to each convolution kernel; A determination module, configured to determine a first set of images according to all the subsets of images; A blood flow image calculation module, configured to pair the images in the subset of images corresponding to each convolution kernel based on a preset pairing rule to obtain image pairs; for each of the image pairs, perform decorrelation calculation on the image pair to obtain a first intermediate result corresponding to the image pair; determine a first target result corresponding to each convolution kernel according to all the first intermediate results; calculate a blood flow image according to all the first target results.

9. An electronic device, characterized in that, Comprising: A memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to implement the method according to any one of claims 1-7.

Images