Film phase unwinding method, device, electronic device and storage medium

Abstract

The embodiment of the present invention discloses a method, device, electronic device and storage medium for movie phase unwrapping. The method comprises: collecting multiple images to be processed of a target object within a preset time period, wherein the multiple images to be processed correspond to different cardiac motion phases and the images to be processed have entangled phases; inputting the multiple images to be processed into a trained unwrapping model to obtain multiple corrected images, wherein the corrected images are obtained by unwrapping the entangled phases of the images to be processed, and the unwrapping model is trained based on multiple historical sample data with paired entangled phases and historical sample data with suppressed entangled phases. This achieves the effect of efficiently and accurately unwrapping the movie phase.

Description

Technical Field

[0001] The embodiments of the present invention relate to image processing technology, and in particular to a film phase unwrapping method, device, electronic device and storage medium. Background Art

[0002] In existing medical superconducting magnetic resonance imaging systems, flow velocity is usually measured through a phase encoding sequence. By applying an encoding gradient along the velocity encoding direction, the phase is proportional to the velocity, and the velocity information is inferred by the phase. However, because the magnetic resonance system can only measure a sequence from -π to π, this limits the maximum flow velocity that can be measured. The encoded flow velocity will affect the sensitivity of the measurable flow velocity. Therefore, it is often necessary to make a balance between the maximum encoded flow velocity and the flow velocity sensitivity (velocity signal-to-noise ratio). When the velocity encoding set in the region of interest is relatively small, phase entanglement will occur in places with relatively large flow velocities outside the region of interest, affecting the accuracy of the results and the recognition effect.

[0003] In the prior art, film phase dewrapping is usually performed by processing each frame of the image separately, derivatizing the dewrapping along the spatial direction, and obtaining dewrapped sequence image information. This dewrapping efficiency is low and the temporal correlation information between frames is not fully utilized. Summary of the invention

[0004] The embodiments of the present invention provide a film phase unwrapping method, device, electronic device and storage medium to achieve the effect of unwrapping the film phase efficiently and accurately.

[0005] In a first aspect, an embodiment of the present invention provides a film phase unwrapping method, the method comprising:

[0006] Collecting a plurality of images to be processed of the target object within a preset time period, wherein the plurality of images to be processed correspond to different cardiac motion phases, and the images to be processed have winding phases;

[0007] Input multiple images to be processed into a trained unwrapping model to obtain multiple corrected images, wherein the corrected images are obtained by unwrapping the unwrapped phases of the images to be processed, and the unwrapping model is trained based on multiple pairs of historical sample data with unwrapped phases and historical sample data with suppressed unwrapped phases.

[0008] In a second aspect, an embodiment of the present invention further provides a film phase unwrapping device, the device comprising:

[0009] The to-be-processed image acquisition module is used to acquire a plurality of to-be-processed images of the target object within a preset time period, wherein the plurality of to-be-processed images correspond to different cardiac motion phases, and the to-be-processed images have winding phases;

[0010] A sequence image information acquisition module is used to input multiple images to be processed into a trained unwrapping model to obtain multiple corrected images, wherein the corrected images are obtained by unwrapping the wrapped phases of the images to be processed, and the unwrapping model is trained based on multiple historical sample data with paired wrapped phases and historical sample data with suppressed wrapped phases.

[0011] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0012] one or more processors;

[0013] A storage device for storing one or more programs;

[0014] When the one or more programs are executed by the one or more processors, the one or more processors implement any of the film phase unwrapping methods described in the embodiments of the present invention.

[0015] In a fourth aspect, an embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to perform any of the film phase unwrapping methods described in the embodiments of the present invention.

[0016] The technical solution of the embodiment of the present invention collects multiple images to be processed of the target object within a preset time period, wherein the multiple images to be processed correspond to different cardiac motion phases; the multiple images to be processed are input into a trained de-wrapping model to obtain multiple corrected images, so that the film phase can be de-wrapped efficiently and accurately. This solves the problem of low de-wrapping accuracy caused by the prior art that only spatial dimension information is relied on to de-wrapped images to obtain de-wrapped sequence image information. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of the film phase unwrapping method in the first embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of the timing of collecting the preview movie and the review movie in the first embodiment of the present invention;

[0019] Figure 3 is a schematic diagram of a sampling trajectory in Embodiment 1 of the present invention;

[0020] Figure 4 is a flow chart of a film phase unwrapping method in Embodiment 2 of the present invention;

[0021] Figure 5 is a schematic diagram of the unwinding process in the second embodiment of the present invention;

[0022] Figure 6 This is a flow chart of a film phase unwrapping method according to a third embodiment of the present invention;

[0023] Figure 7 A schematic diagram of segmenting a region of interest and performing phase unwrapping according to Embodiment 3 of the present invention;

[0024] Figure 8 is a schematic structural diagram of a film phase unwrapping device in Embodiment 3 of the present invention;

[0025] Fig. 9 It is a structural schematic diagram of an electronic device in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0026] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0027] Embodiment 1

[0028] Figure 1 This is a flow chart of cine phase unwrapping provided in Embodiment 1 of the present invention. This embodiment is applicable to the case of unwrapping images of different cardiac motion cycles with time correlation. The method can be executed by a cine phase unwrapping device. The cine phase unwrapping device can be implemented by software and / or hardware. The cine phase unwrapping device can be configured on a computing electronic device, and specifically includes the following steps:

[0029] S110, collecting a plurality of to-be-processed images of the target object within a preset time period, wherein the plurality of to-be-processed images correspond to different cardiac motion phases, and the to-be-processed images have winding phases.

[0030] Exemplarily, the target object may be an object for which multiple images to be processed need to be collected, for example, a person or an animal, etc. The preset time period may be a pre-set time period, for example, 5 minutes, 3-10 minutes, etc.

[0031] The image to be processed may be an image to be dewrapped. The multiple images to be processed correspond to different cardiac motion phases of the target object, and pixels in the image to be processed have entangled phases.

[0032] Optionally, the plurality of images to be processed are obtained by a prospective movie or a retrospective movie method.

[0033] For example, Figure 2The schematic diagram of the timing of prospective and retrospective movie acquisition. Since the number of phases of the generated cardiac cycle image is related to the patient's cardiac cycle and the required time resolution of the sequence, the number of cardiac phases will vary dynamically from person to person.

[0034] In the prospective movie acquisition and reconstruction method, the physician will calculate the M cardiac cycles required in advance by estimating the cardiac cycles in advance, and then issue M cardiac cycle imaging sequence segments. The final reconstruction is also M phases, where M is a positive integer.

[0035] The process of retrospective movie acquisition is to start acquiring at the beginning of the R wave of the cardiac cycle and stop acquiring until the beginning of the R wave of the next cardiac cycle. The number of phases acquired may be M or N. Finally, the required number of phases is reconstructed through interpolation during reconstruction, where N is a positive integer.

[0036] Image data of multiple phases can be acquired in the following manner: first, a sampling trajectory is set, the sampling pattern includes multiple sampling points, each sampling point represents a k-space line associated with k-space data, and the sampling points are set so that adjacent sampling points along the time axis correspond to different phase encoding gradient field values; a scanning sequence is excited according to the sampling trajectory to obtain k-space data; and the k-space data is reconstructed to obtain multiple images.

[0037] Figure 3 FIG. 1 is a schematic diagram showing a sampling trajectory according to an embodiment of the present invention. Figure 3 As shown in FIG. 1 , the horizontal axis represents the time dimension. The vertical axis represents the phase encoding dimension of the phase encoding gradient field value. Figure 3 The black dots shown indicate sampling points with corresponding specific time values ​​and specific phase encoding gradient field values. The figure contains multiple sampling points, which correspond to images of 6 phases. In this embodiment, the image of each phase corresponds to five sampling points. The images of any two adjacent phases have sampling trajectories in different directions, and the adjacent phases share a sampling point. Taking phase 1, phase 2 and phase 3 as examples, phase 1 and phase 3 are both adjacent phases of phase 2. The phase encoding gradient field values ​​corresponding to the sampling points of phase 1 change in descending order along the time dimension, so that the sampling trajectory relative to phase 1 has a descending direction (indicated by Figure 3 The phase encoding gradient field values ​​corresponding to the sampling points of phase 2 change in ascending order along the time dimension, so that the sampling trajectory relative to phase 2 has an ascending direction (indicated by the arrow A in FIG. 1 ). Figure 3Indicated by arrow B in FIG. 1 ). Similar to phase 1, the sampling trajectory relative to phase 3 has a descending direction. Phase 1 and phase 2 share a sampling point at the edge of the sampling trajectory (here is the minimum position of the phase encoding gradient field value). Phase 2 and phase 3 share a sampling point at the edge of the sampling trajectory (here is the maximum position of the phase encoding gradient field value). Using the method of the embodiment of the present application, discontinuous sampling is performed along the phase encoding dimension, saving scanning time and avoiding patients from holding their breath for a long time; the images of two adjacent phases have sampling trajectories in different directions, and adjacent phases share a sampling point, which can be applicable to patients with irregular heart rates.

[0038] It should be noted that the cardiac motion phases here can be different cardiac motion phases in a cardiac cycle of the target object. The cardiac cycle here can be: P wave, QRS complex wave, T wave, ST segment, forming a cardiac cycle, P wave indicates that the heart enters the atrial contraction period, Q wave indicates that the heart is about to enter the isovolumetric contraction period, S wave indicates that the heart is in the isovolumetric contraction period, S wave to T wave indicates the stage of transition from the contraction period to the diastole period, and R wave indicates that the heart enters the isovolumetric contraction period from the atrial contraction period. There can be multiple cardiac motion phases in each cardiac cycle.

[0039] In this way, a plurality of dewrapped images can be obtained subsequently based on obtaining a plurality of to-be-processed images of the target object within a preset time period.

[0040] In this embodiment, an ECG waveform collection device is used to collect the ECG waveform of the time-varying electrical phenomenon of the heart of the target object. In the scanning stage, the R wave generation time phase in the ECG waveform is used as a trigger, and the time interval between the current R wave and the next R wave is a cardiac cycle. In one embodiment, a cardiac cycle includes a cardiac systolic period and a cardiac diastolic period, wherein the cardiac systolic period includes atrial systolic period, isovolumetric contraction period, rapid ejection period, and slow ejection period; the cardiac diastolic period includes multiple phases such as isovolumetric relaxation period, rapid filling period, slow filling period, and late diastolic period.

[0041] S120. Input a plurality of images to be processed into a trained unwrapping model to obtain a plurality of corrected images, wherein the corrected images are obtained by unwrapping the unwrapped phases of the images to be processed, and the unwrapping model is trained based on a plurality of historical sample data of paired unwrapped phases and historical sample data of suppressed unwrapped phases.

[0042] Exemplarily, the corrected image may be an image output by a de-wrapping model after the image to be processed is input into a trained de-wrapping model, and specifically may be an image obtained by de-wrapping the de-wrapped phase of the image to be processed.

[0043] It should be noted that the calibrated image here may be a plurality of heart images.

[0044] The disentanglement model may be a model that can directly obtain sequence image information of multiple disentangled images after multiple images to be processed are input into the model. For example, the disentanglement model may be a neural network model, a decision tree model, a support vector machine model, etc.

[0045] It should be noted that the disentanglement model here can be obtained by training based on a plurality of historical sample data of paired entangled phases and historical sample data of suppressed entangled phases.

[0046] When a plurality of images to be processed are input into the de-entanglement model, the de-entanglement model can distinguish the sequence information of the same blood flow velocity in the plurality of images to be processed based on the input plurality of images to be processed.

[0047] By inputting multiple images to be processed into the trained de-wrapping model, the sequential image information of multiple de-wrapped images can be obtained, so that the film phase can be de-wrapped efficiently and accurately. This solves the problem of low de-wrapping accuracy caused by the prior art that only relies on spatial dimension information to de-wrapped the images to be processed, and then obtains the de-wrapped sequential image information.

[0048] The technical solution of the embodiment of the present invention collects multiple images to be processed of the target object within a preset time period, wherein the multiple images to be processed correspond to different cardiac motion phases; the multiple images to be processed are input into a trained de-wrapping model to obtain multiple corrected images, so that the film phase can be de-wrapped efficiently and accurately. This solves the problem of low de-wrapping accuracy caused by the prior art that only spatial dimension information is relied on to de-wrapped images to obtain de-wrapped sequence image information.

[0049] Embodiment 2

[0050] Figure 4 A flow chart of a film phase unwrapping method provided for Embodiment 2 of the present invention, the embodiment of the present invention can be combined with the various optional schemes in the above embodiments. In the embodiment of the present invention, optionally, before inputting a plurality of images to be processed into a trained unwrapping model, the method further includes: determining a first number set at the input end of the unwrapping model; comparing the first number with a second number of collected images to be processed, and preprocessing the images to be processed based on the comparison result, wherein the number of images to be processed after preprocessing satisfies the first number. Inputting a plurality of images to be processed into a trained unwrapping model to obtain a plurality of corrected images includes: inputting a first number of preprocessed images to be processed into the unwrapping model to obtain a first number of intermediate corrected images; and processing the first number of sequence image information to obtain a second number of corrected images.

[0051] like Figure 4 As shown, the method of the embodiment of the present invention specifically includes the following steps:

[0052] S210, collecting a plurality of to-be-processed images of the target object within a preset time period, wherein the plurality of to-be-processed images correspond to different cardiac motion phases, and the to-be-processed images have winding phases.

[0053] S220: Determine a first quantity set at an input end of the detangling model.

[0054] Exemplarily, the first number may be the number of images input into the dewrapping model set at the input end.

[0055] Due to the different number of phases of cardiac motion, the dimension of the data structure input into the de-wrapping model is also different, which affects the de-wrapping of multiple images to be processed. Therefore, before inputting multiple images to be processed into the de-wrapping model, the number of images to be processed should be set to the number of images to be input into the de-wrapping model. The specific example is as follows: Figure 5 The schematic diagram of the unwinding process is shown in Figure 5 In the figure, the middle area A is the disentanglement model. In the disentanglement model, each box represents a different structural layer. Specifically, box M is a convolution layer + a normalization layer + a linear rectification layer, box N is a full connection layer, box H is an upsampling layer, and box K is a maximum transfer function layer.

[0056] Specifically, the number of images to be processed is set to the number of images input into the dewrapping model as specified by the dewrapping model, which can be as follows: Figure 5 As shown, a preprocessing layer is added before the input of the disentanglement model to process the number of images to be processed into the number of images consistent with the input of the disentanglement model. Specifically, linear interpolation can be performed along the time direction to maintain the consistency of the size of the corresponding disentanglement model input matrix.

[0057] Before processing the number of images to be processed to be consistent with the number of images at the input end of the dewrapping model, it is necessary to first determine the first number set at the input end of the dewrapping model, so that the acquired multiple images to be processed can be processed based on the first number.

[0058] The interpolation method of the present application can be generally applied to images obtained by both retrospective and prospective methods, without the need to set up a network for each method, and is more versatile.

[0059] S230, comparing the first number with the second number of collected images to be processed, and preprocessing the images to be processed based on the comparison result, wherein the number of the images to be processed after the preprocessing meets the first number.

[0060] Exemplarily, the second number may be the number of acquired images to be processed. Optionally, the second number here is less than or equal to the first number.

[0061] After obtaining the number of images input into the de-entanglement model, that is, the first number, the number of acquired images to be processed is compared with the first number. When the number of acquired images to be processed is inconsistent with the number of images input into the de-entanglement model, the images to be processed need to be preprocessed so that the number of images to be processed after preprocessing is consistent with the number of images input into the de-entanglement model.

[0062] Optionally, preprocessing the image to be processed based on the comparison result may specifically be: when the first number is greater than the second number, interpolating the image to be processed based on the first number to obtain interpolated image to be processed, so that the number of interpolated images to be processed is equal to the first number.

[0063] Exemplarily, the interpolated image to be processed may be an image obtained by interpolating the originally acquired image to be processed.

[0064] After obtaining the number of images input into the de-entanglement model specified by the model, i.e., the first number, and the number of images to be processed originally acquired, i.e., the second number, the first number is compared with the second number. If the first number is greater than the second number, the second number of images to be processed are interpolated based on the first number to obtain interpolated images to be processed, so that the number of images to be processed after interpolation is equal to the first number.

[0065] For example, the first number is 30, the cardiac cycle of the originally acquired images to be processed is 1200ms, and the number of the originally acquired images to be processed is 3, which are acquired at 0ms, 600ms, and 1200ms. Since 30>3, the number of the originally acquired images to be processed needs to be interpolated, and 27 images need to be interpolated, so 27 images are interpolated images to be processed, and the number of images to be processed after interpolation is 30.

[0066] After comparing the first number with the second number, if the first number is smaller than the second number, the second number of images to be processed are deleted based on the first number. Specifically, the difference between the original acquired multiple images to be processed and the first number is calculated, and then the dewrapped images of the calculated difference number are deleted from the original acquired multiple images to be processed, so that the number of images to be processed after deletion is equal to the first number.

[0067] For example, the first number is 30, the cardiac cycle of the originally acquired images to be processed is 1200ms, and the number of the originally acquired images to be processed is 33. Since 33>30, the number of the originally acquired images to be processed is 3 more than the first number. Then, 3 unwrapped images are deleted from the originally acquired images to be processed, so that the number of images to be processed after deletion is 30.

[0068] In this way, according to the comparison result of the first number and the second number, the second number of images to be processed are interpolated or deleted, thereby ensuring the consistency of the number of images to be processed in the input dewrapping model.

[0069] Optionally, based on the first quantity, interpolating the image to be processed to obtain the interpolated image to be processed may specifically be: determining a third quantity of the interpolated images to be processed based on the first quantity and the second quantity; determining a first weight of each image to be processed to be interpolated based on the timestamp of any interpolated image to be processed and the timestamp of the image to be processed to be interpolated, wherein the first weight is determined based on the timestamp of the current interpolated image to be processed and the difference between the timestamps of two images to be processed to be interpolated that are adjacent to the timestamp of the current interpolated image to be processed; interpolation is performed based on the image to be processed to be interpolated and the corresponding first weight to obtain the interpolated image to be processed.

[0070] Exemplarily, the third number may be the number of dewrapped images that need to be interpolated calculated based on the first number and the second number. For example, if the first number is 30 and the number of originally acquired images to be processed is 3, then 27 of the originally acquired images to be processed need to be interpolated, and these 27 images are the third number.

[0071] The timestamp of the interpolated image to be processed may be the time when the image to be processed is interpolated. The timestamp of the interpolated image to be processed may be the acquisition time of the interpolated image to be processed.

[0072] The first weight may be a weight of the image to be processed that is interpolated and determined based on the timestamp of the interpolated image to be processed and the timestamp of the image to be processed that is interpolated. Specifically, the first weight is determined based on the timestamp of the current interpolated image to be processed and the difference between the timestamps of two images to be processed that are adjacent to the timestamp of the current interpolated image to be processed.

[0073] For example, if the timestamp of the current interpolation image to be processed is 40ms, and the timestamps of the two interpolation images to be processed adjacent to the timestamp of the current interpolation image to be processed are 0ms and 600ms, respectively, then the weights of the interpolation image to be processed acquired at 0ms and the interpolation image to be processed acquired at 600ms are determined based on the difference between 40ms and 0ms and 600ms, respectively. For example, the difference between 40ms and 0ms is smaller than the difference between 40ms and 600ms, so the interpolation image to be processed acquired at 0ms can be assigned a larger weight, and the interpolation image to be processed acquired at 600ms can be assigned a smaller weight.

[0074] After determining the third number of interpolated images to be processed, the first weight of the image to be processed to be interpolated can be determined according to the timestamp of any interpolated image to be processed and the timestamp of the image to be processed to be interpolated, and the interpolated image to be processed can be obtained according to the image to be processed to be interpolated and the first weight corresponding to the image to be processed to be interpolated.

[0075] For example, the first number is 30, the cardiac cycle of the originally acquired images to be processed is 1200ms, the number of the originally acquired images to be processed is 3, which are acquired at 0ms, 600ms and 1200ms, and the third number of interpolated images to be processed is 27. Then the 3 originally acquired images to be processed are the images to be processed. Since the first number is 30, and the cardiac cycle of the originally acquired images to be processed is 1200ms, images should be acquired at 0ms, 40ms, 80ms, 120ms, ..., 1200ms, respectively. Since the number of the originally acquired images to be processed is 3, which are acquired at 0ms, 600ms and 1200ms, images should be interpolated at 40ms, 80ms, 120ms, 560ms, 640ms, ..., 1160ms on the basis of the images to be processed, and the interpolated images at 40ms, 80ms, 120ms, 560ms, 640ms, ..., 1160ms are the interpolated images to be processed. According to the above-mentioned first weight allocation method, images are interpolated at 40ms, 80ms, 120ms, 560ms, 640ms, ..., 1160ms to obtain interpolated images to be processed.

[0076] In this way, based on the timestamp of any interpolated image to be processed and the timestamp of the image to be processed undergoing interpolation processing, the first weight of each image to be processed undergoing interpolation processing is determined, and interpolation processing is performed according to the image to be processed undergoing interpolation processing and the corresponding first weight to obtain the interpolated image to be processed. In this way, the interpolated image to be processed can be closer to the originally collected image to be processed, thereby ensuring the authenticity of the image to be processed input into the de-wrapping model, so that the obtained sequence image information of the de-wrapped image is also more accurate and authentic.

[0077] S240: Input the preprocessed first number of to-be-processed images into a dewrapping model to obtain a first number of intermediate corrected images, and process the first number of intermediate corrected images to obtain a second number of corrected images.

[0078] Exemplarily, the intermediate corrected images may be obtained by inputting a first number of pre-processed images to be processed into a dewrapping model to obtain a first number of dewrapped images.

[0079] A first number of pre-processed images to be processed are input into the dewrapping model to obtain a first number of intermediate corrected images.

[0080] However, since the first number of images to be processed are preprocessed before passing through the dewrapping model, part of the information in the first number of intermediate corrected images is obtained by interpolating the images to be processed. Therefore, a processing module can be added before the output layer of the dewrapping model to adjust the first number of intermediate corrected images into corrected images of the second number of images to be processed originally acquired.

[0081] For example, the number of originally acquired images to be processed is 3. Before inputting the de-wrapping model, the originally acquired de-wrapped images are preprocessed to obtain 30 images to be processed as specified by the de-wrapping model. These 30 images to be processed are input into the de-wrapping model to obtain intermediate corrected images of the 30 images to be processed. After obtaining the intermediate corrected images of the 30 images to be processed, the intermediate corrected images of the 30 images to be processed are processed so that the corrected images input from the de-wrapping model are the number of originally acquired images to be processed, that is, 3 corrected images.

[0082] Optionally, the processing of the first number of sequence image information may specifically include: deleting the sequence image information corresponding to the interpolated image to be processed obtained by interpolation processing.

[0083] Exemplarily, if the image to be processed in the input dewrapping model is obtained by interpolation, the intermediate correction image corresponding to the interpolated image to be processed obtained by interpolation is deleted. For example, the first number is 30, the number of originally collected images to be processed is 3, and the number of interpolated images to be processed is 27. After obtaining 30 intermediate correction images, the intermediate correction images corresponding to the 27 interpolated images to be processed obtained by interpolation are deleted.

[0084] In this way, it is ensured that the number of the sequence image information of the output images to be processed is equal to the number of the originally acquired images to be processed, thereby ensuring the consistency of the data structure of the de-wrapping model.

[0085] The technical solution of the embodiment of the present invention determines the first number set at the input end of the dewrapping model, compares the first number with the second number of images to be processed collected, and pre-processes the images to be processed based on the comparison result, so that the second number of images to be processed are interpolated according to the comparison result of the first number and the second number, thereby ensuring the consistency of the number of images to be processed in the input dewrapping model. The first number of images to be processed after preprocessing are input into the dewrapping model to obtain the first number of intermediate corrected images, and the first number of intermediate corrected images are processed to obtain the second number of corrected images, thereby ensuring that the number of corrected images of the output images to be processed is equal to the number of images to be processed originally collected, thereby ensuring the consistency of the data structure of the dewrapping model.

[0086] Embodiment 3

[0087] Figure 6 This is a flow chart of a film phase unwrapping method provided in Embodiment 3 of the present invention. The embodiment of the present invention can be combined with each optional solution in the above embodiments. In the embodiment of the present invention, optionally, if there is a region of interest in the image to be processed and phase unwrapping occurs in the region of interest, it is only necessary to perform phase unwrapping on the region of interest where phase unwrapping occurs.

[0088] like Figure 6 As shown, the method of the embodiment of the present invention specifically includes the following steps:

[0089] S310, collecting a plurality of images to be processed of the target object within a preset time period, wherein the plurality of images to be processed correspond to different cardiac motion phases, and the images to be processed have regions of interest with entangled phases.

[0090] Exemplarily, the image to be processed is a phase image. The phase image can be a two-dimensional phase image, or a three-dimensional phase image or other multi-dimensional phase images above three dimensions. For a two-dimensional phase image, one dimension corresponds to the phase encoding direction, and the other dimension corresponds to the frequency encoding direction. For a three-dimensional phase image, the first dimension corresponds to the phase encoding direction, the second dimension corresponds to the frequency encoding direction, and the third dimension may correspond to the layer selection direction. Of course, the phase image may also include a time dimension.

[0091] The region of interest here may be a region formed by a plurality of pixels having twisted phases in the image to be processed.

[0092] S320: Determine a first quantity set at an input end of the detangling model.

[0093] S330, comparing the first number with the second number of collected images to be processed, and preprocessing the images to be processed based on the comparison result, wherein the number of the images to be processed after the preprocessing meets the first number.

[0094] S340: Input the preprocessed first number of to-be-processed images to the de-wrapping model, so that the de-wrapping model processes the to-be-processed images, segments the region-of-interest images, and outputs a first number of corrected region-of-interest images.

[0095] S350: Process the first number of corrected region-of-interest images to obtain a second number of corrected region-of-interest images.

[0096] Exemplarily, after the second number of corrected region-of-interest images are obtained, the pixels of the region of interest included in the image to be processed are replaced with the corrected region-of-interest images to obtain the final corrected image.

[0097] Exemplarily, a plurality of pixels having a wrapped phase in the corrected region of interest image are corrected.

[0098] In the embodiment of the present application, only the image of the region of interest with the winding phase is processed, which can significantly improve the image processing speed.

[0099] For details, please refer to Figure 7 The schematic diagram of segmenting the region of interest for phase unwrapping is as follows: Figure 7The figure shows seven images obtained by scanning the target object, each of which corresponds to a phase of the target object's physiological movement. In the figure, the images obtained by scanning the six phases 1-3, 5-7 have no wrapping phases, while the image obtained by scanning the fourth phase has wrapping phases. The region of interest (circled area in the figure) can be determined in the image scanned in the fourth phase by taking the derivative along the spatial direction. Using the same method, the parts corresponding to the region of interest are also segmented in the images scanned in other phases. Next, the region of interest corresponding to the fourth phase and the parts corresponding to the region of interest in other phases are obtained using the trained disentanglement model to obtain a corrected region of interest image.

[0100] Specifically, since the phase information is obtained by calculating the inverse trigonometric function, the storage value range of the phase information is only 2π, generally (-π,π] or (0,2π]. If the actual phase value exceeds this range, the stored value of the phase will be folded into the storage range, resulting in an error of 2nπ (n is a non-zero integer), forming a jump artifact on the image. To restore the true phase value, phase unwrapping calculation is required.

[0101] After obtaining the image to be processed, it is necessary to preprocess the image to be processed to obtain a phase mask corresponding to the image to be processed. Specifically, it can be: assuming that the phase map is a two-dimensional phase map Q, the phase data of each point in the two-dimensional phase map Q is shown in the matrix A of the following formula (1).

[0102]

[0103] The row direction of matrix A is along the X-axis or phase encoding direction, and the column direction is along the Y-axis or frequency encoding direction. The data in matrix A represents the phase value of each point in the two-dimensional phase diagram. Among them, the phase mask includes two points (x2, y2) and (x3, y3), and the data corresponding to the point (x2, y2) is a 22 , the data corresponding to the point (x3, y3) is a 23 .

[0104] Secondly, according to the phase mask, the image to be processed is segmented along the first direction and the second direction respectively to obtain at least two segmented phase images.

[0105] In this embodiment, the image to be processed for each phase is segmented at least along the first direction and the second direction, wherein the first direction may be the phase encoding direction and the second direction may be the frequency encoding direction. In this embodiment, according to the phase mask, the two-dimensional phase image Q is segmented along the X-axis direction to obtain five one-dimensional data segments in the X-direction, which are [a 11 ,a 12 ,a 13 ,a 14]、[a 21 ]、[a 24 ]、[a 31 ,a 32 ,a 33 ,a 34 ]、[a 41 ,a 42 ,a 43 ,a 44 ], these five one-dimensional data segments in the X direction constitute the segmented phase image in the X-axis direction. According to the phase mask, the two-dimensional phase image Q is segmented along the Y-axis direction to obtain six one-dimensional data segments in the Y direction, as shown below:

[0106] [a 12 ]、 [a 13 ]、

[0107] Similarly, these six one-dimensional data segments in the Y direction constitute a segmented phase map in the Y-axis direction.

[0108] Again, the trained disentanglement model is used to process the segmented phase diagrams of multiple phases to obtain the corrected segmented phase diagrams along the X-axis and Y-axis directions. Correspondingly, the trained disentanglement model also selects multiple historical segmented phase diagrams along the X-axis direction and the historical segmented phase diagrams along the Y-axis direction during the training stage. The historical segmented phase diagrams along the X-axis and Y-axis directions respectively include multiple pairs of historical segmented phase diagrams with entangled phases and historical segmented phase diagrams without entangled phases. Among them, the historical segmented phase diagram with entangled phases is obtained by processing the historical sample data with entangled phases as described above.

[0109] Finally, the corrected segmented phase images along the X and Y directions belonging to the same phase are merged to obtain the corrected image of each phase.

[0110] The technical solution of the embodiment of the present application can improve the stability of image correction by segmenting the phase image along at least two directions and then performing phase unwrapping processing on at least two segmented phase images obtained by segmentation.

[0111] Embodiment 4

[0112] Figure 8 This is a schematic diagram of the structure of a film phase unwrapping device provided in Embodiment 4 of the present invention. Figure 8 As shown, the device includes: a to-be-processed image acquisition module 31 and a corrected image acquisition module 32.

[0113] The to-be-processed image acquisition module 31 is used to acquire a plurality of to-be-processed images of the target object within a preset time period, wherein the plurality of to-be-processed images correspond to different cardiac motion phases, and the to-be-processed images have winding phases;

[0114] The corrected image acquisition module 32 is used to input the multiple images to be processed into a trained de-wrapping model to obtain multiple corrected images, wherein the de-wrapping model is trained based on multiple pairs of historical sample data with wrapped phases and historical sample data with suppressed wrapped phases.

[0115] Based on the technical solution of the above embodiment, the device further includes:

[0116] A first quantity acquisition module, used to determine a first quantity set at an input end of a detangling model;

[0117] A preprocessing module is used to compare the first number with a second number of the collected images to be processed, and preprocess the images to be processed based on the comparison result, wherein the number of the preprocessed images to be processed meets the first number.

[0118] Based on the technical solution of the above embodiment, the preprocessing module includes:

[0119] The first preprocessing unit is used to interpolate the images to be processed based on the first number to obtain interpolated images to be processed when the first number is greater than the second number, so that the number of the images to be processed after interpolation is equal to the first number.

[0120] Based on the technical solution of the above embodiment, the first preprocessing unit is specifically used for:

[0121] When the first number is greater than the second number, determining a third number of interpolated images to be processed based on the first number and the second number;

[0122] Determine, based on the timestamp of any interpolated image to be processed and the timestamp of the image to be processed to be interpolated, a first weight of each of the images to be processed to be interpolated, wherein the first weight is determined based on the timestamp of the current interpolated image to be processed and the difference between the timestamps of two images to be processed to be interpolated that are adjacent to the timestamp of the current interpolated image to be processed;

[0123] An interpolation process is performed based on the image to be processed and the corresponding first weight to obtain the interpolation image to be processed.

[0124] Based on the technical solution of the above embodiment, the corrected image acquisition module 32 includes:

[0125] An intermediate correction image acquisition unit, used for inputting a first number of pre-processed images to be processed into the dewrapping model to obtain a first number of intermediate correction images;

[0126] The corrected image acquisition unit is used to process the first number of intermediate corrected images to obtain a second number of corrected images.

[0127] Based on the technical solution of the above embodiment, the correction image acquisition unit includes:

[0128] A corrected image acquisition subunit, used for deleting the intermediate corrected images corresponding to the interpolated images to be processed obtained by the interpolation process, to obtain a second number of intermediate corrected images;

[0129] The film phase unwrapping device provided in the embodiment of the present invention can execute the film phase unwrapping method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0130] Embodiment 5

[0131] Fig. 9 A schematic diagram of the structure of an electronic device provided in Embodiment 5 of the present invention is shown in FIG. Fig. 9 As shown, the electronic device includes a processor 70, a memory 71, an input device 72 and an output device 73; the number of processors 70 in the electronic device can be one or more. Fig. 9 A processor 70 is taken as an example; the processor 70, the memory 71, the input device 72 and the output device 73 in the electronic device can be connected by a bus or other means. Fig. 9 The example of connecting through bus is taken in the following.

[0132] The memory 71, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the film phase unwrapping method in the embodiment of the present invention (for example, the to-be-processed image acquisition module 31 and the corrected image acquisition module 32). The processor 70 executes various functional applications and data processing of the electronic device by running the software programs, instructions and modules stored in the memory 71, that is, implements the above-mentioned film phase unwrapping method.

[0133] The memory 71 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 71 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 71 may further include a memory remotely arranged relative to the processor 70, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0134] The input device 72 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the electronic device. The output device 73 may include a display device such as a display screen.

[0135] Embodiment 6

[0136] Embodiment 6 of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform a film phase unwrapping method when executed by a computer processor.

[0137] Of course, the computer executable instructions of a storage medium containing computer executable instructions provided in an embodiment of the present invention are not limited to the operations of the method described above, and can also execute related operations in the film phase unwrapping method provided in any embodiment of the present invention.

[0138] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer electronic device (which can be a personal computer, a server, or a network electronic device, etc.) to execute the methods described in each embodiment of the present invention.

[0139] It is worth noting that in the embodiment of the above-mentioned film phase unwrapping device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0140] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A film phase unwrapping method, characterized in that: include: Acquire multiple images to be processed of the target object within a preset time period, wherein the multiple images to be processed correspond to different cardiac motion phases, and the images to be processed have winding phases; Inputting a plurality of the images to be processed into a trained unwrapping model to obtain a plurality of corrected images, wherein the corrected images are obtained by unwrapping the unwrapped phases of the images to be processed, and the unwrapping model is trained based on a plurality of historical sample data with paired unwrapped phases and historical sample data with suppressed unwrapped phases; Wherein, before inputting the plurality of images to be processed into the trained de-entanglement model, the method further includes: Determining a first quantity set at an input end of the unwrapping model; Comparing the first number with a second number of the collected images to be processed, and preprocessing the images to be processed based on the comparison result, wherein the number of the preprocessed images to be processed meets the first number; The step of inputting the plurality of images to be processed into a trained de-wrapping model to obtain a plurality of corrected images comprises: Inputting a first number of preprocessed images to be processed into the dewrapping model to obtain a first number of intermediate corrected images; Processing the first number of intermediate corrected images to obtain a second number of corrected images; The processing of the first number of intermediate corrected images includes: The intermediate corrected image corresponding to the interpolated image to be processed obtained by the interpolation process is deleted.

2. The method according to claim 1, characterized in that: The preprocessing of the image to be processed based on the comparison result includes: When the first number is greater than the second number, the images to be processed are interpolated based on the first number to obtain interpolated images to be processed, so that the number of the images to be processed after interpolation is equal to the first number.

3. The method according to claim 2, characterized in that The interpolating the image to be processed based on the first quantity to obtain an interpolated image to be processed includes: Determine a third number of interpolated images to be processed based on the first number and the second number; Determine, based on the timestamp of any interpolated image to be processed and the timestamp of the image to be processed to be interpolated, a first weight of each of the images to be processed to be interpolated, wherein the first weight is determined based on the timestamp of the current interpolated image to be processed and the difference between the timestamps of two images to be processed to be interpolated that are adjacent to the timestamp of the current interpolated image to be processed; An interpolation process is performed based on the image to be processed and the corresponding first weight to obtain the interpolation image to be processed.

4. The method according to claim 1, characterized in that The plurality of images to be processed are obtained by a prospective movie method or a retrospective movie method.

5. A film phase unwrapping device, characterized in that: include: The to-be-processed image acquisition module is used to acquire a plurality of to-be-processed images of the target object within a preset time period, wherein the plurality of to-be-processed images correspond to different cardiac motion phases, and the to-be-processed images have winding phases; A correction image acquisition module, used for inputting a plurality of the images to be processed into a trained unwrapping model to obtain a plurality of correction images, wherein the correction images are obtained by unwrapping the unwrapped phases of the images to be processed, and the unwrapping model is trained based on a plurality of historical sample data with paired unwrapped phases and historical sample data with suppressed unwrapped phases; Wherein, the film phase unwrapping device further comprises: A first quantity acquisition module, used to determine a first quantity set at the input end of the detangling model; a preprocessing module, configured to compare the first number with a second number of the collected images to be processed, and preprocess the images to be processed based on the comparison result, wherein the number of the preprocessed images to be processed meets the first number; Wherein, the corrected image acquisition module comprises: An intermediate correction image acquisition unit, used for inputting a first number of pre-processed images to be processed into the dewrapping model to obtain a first number of intermediate correction images; a corrected image acquisition unit, configured to process the first number of intermediate corrected images to obtain a second number of corrected images; Wherein, the correction image acquisition unit comprises: The corrected image acquisition subunit is used to delete the intermediate corrected images corresponding to the interpolated images to be processed obtained by the interpolation process to obtain a second number of intermediate corrected images.

6. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the film phase unwrapping method as described in any one of claims 1-4.

7. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to perform the film phase unwrapping method as claimed in any one of claims 1 to 4 when executed by a computer processor.

Images