A two-dimensional brain imaging full waveform inversion method, device, medium and product

Abstract

This invention discloses a method, apparatus, medium, and product for full-waveform inversion of two-dimensional cranial imaging, relating to the field of cranial imaging technology. The method includes: acquiring cranial CT images; the cranial CT images include skull CT images, soft tissue CT images, and vascular CT images; performing an ultrasound experiment on the cranial CT images using a transducer array to obtain acoustic wave information; performing inversion based on an initial predicted signal wave and the acoustic wave information to obtain an inverted image; registering and fusing the inverted image with the cranial CT image as a reference to obtain a complete cranial inverted acoustic velocity map; training a deep learning network using the complete cranial inverted acoustic velocity map as input and the cranial CT image as output to obtain an inversion model; the inversion model is used to invert the complete cranial inverted acoustic velocity map. This invention can reduce the time required for inversion imaging.

Description

Technical Field

[0001] This invention relates to the field of cranial imaging technology, and in particular to a two-dimensional cranial imaging full waveform inversion method, device, medium and product. Background Technology

[0002] Transcranial ultrasound (TCE) suffers from ultrasound propagation path deflection and strong reflections caused by the skull, preventing direct brain imaging. Recently, a full-waveform inversion algorithm based on USCT technology has been applied to brain imaging. Developed in the context of solving high-resolution seismic imaging, the full-waveform inversion algorithm is an optimized method for solving forward modeling problems. It eliminates the need for beamforming, is unaffected by strong reflections from the skull, is easy to transmit, and boasts a good signal-to-noise ratio. Compared to traditional methods in this field, it promises to solve the challenge of reconstructing the porous structure of the skull during imaging. However, the full-waveform inversion algorithm suffers from two major drawbacks: long computation time, which fails to meet the real-time requirements of clinical applications, and its iterative results depend on the initial model. Summary of the Invention

[0003] The purpose of this invention is to provide a method, apparatus, medium, and product for full waveform inversion of two-dimensional cranial imaging, which can reduce the time required for inversion imaging.

[0004] To achieve the above objectives, the present invention provides the following solution:

[0005] A method for full waveform inversion in two-dimensional cranial imaging includes:

[0006] Acquire cranial CT images; the cranial CT images include skull CT images, soft tissue CT images, and vascular CT images;

[0007] An ultrasound experiment was performed on the cranial CT image using a transducer array to obtain sound wave information;

[0008] The inversion image is obtained by performing an inversion based on the initial predicted signal wave and the acoustic wave information;

[0009] The inverted image is registered and fused with the cranial CT image as a reference to obtain a complete cranial inverted sound velocity map.

[0010] Using the complete inverted sound velocity map of the brain as input and the brain CT image as output, a deep learning network is trained to obtain an inversion model; the inversion model is used to invert the complete inverted sound velocity map of the brain.

[0011] Optionally, the initial predicted signal wave is obtained from an ultrasonic emission simulation experiment using a pure water model.

[0012] Optionally, an inversion image is obtained by performing an inversion based on the initial predicted signal wave and the acoustic wave information, specifically including:

[0013] Based on the initial predicted signal wave and the acoustic wave information, an inversion algorithm is used to perform inversion and obtain the inverted image.

[0014] Optionally, based on the initial predicted signal wave and the acoustic wave information, a full waveform inversion algorithm is used to invert the waveform and obtain an inverted image, specifically including:

[0015] Based on the initial predicted value signal wave, the predicted value is obtained by performing forward modeling using the finite element method;

[0016] The inversion image is obtained by iterating using the steepest gradient method based on the predicted value and the acoustic wave information.

[0017] Optionally, the acoustic information includes skull acoustic information, soft tissue acoustic information, and blood vessel acoustic information.

[0018] The present invention also provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method.

[0019] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method.

[0020] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method.

[0021] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0022] This invention utilizes a transducer array to perform ultrasound experiments on the cranial CT image to obtain acoustic wave information; it then performs inversion based on the initial predicted signal wave and the acoustic wave information to obtain an inverted image; the inverted image is registered and fused with the cranial CT image as a reference to obtain a complete cranial inverted acoustic velocity map; using the complete cranial inverted acoustic velocity map as input and the cranial CT image as output, a deep learning network is trained to obtain an inversion model; when processing the cranial CT image, different parts are processed separately to avoid unclear images that lead to indistinguishable tissues; and the inversion model reduces the time required for inversion imaging. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 The full waveform inversion flowchart provided by this invention;

[0025] Figure 2 This is a schematic diagram of the full waveform inversion method for two-dimensional cranial imaging;

[0026] Figure 3 This is a full waveform inversion diagram of brain tissue;

[0027] Figure 4 The flowchart of the two-dimensional cranial imaging full waveform inversion method provided by the present invention is shown. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] The purpose of this invention is to provide a method, apparatus, medium, and product for full waveform inversion of two-dimensional cranial imaging, which can reduce the time required for inversion imaging.

[0030] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] like Figure 2 and Figure 4 As shown, the present invention provides a two-dimensional cranial imaging full waveform inversion method, comprising:

[0032] Step 101: Obtain cranial CT images; the cranial CT images include skull CT images, soft tissue CT images and vascular CT images.

[0033] Experimental data acquisition: Based on the different acoustic parameters in CT images, the cranium was separated into three parts: skull, soft tissue, and blood vessels.

[0034] Step 102: Perform an ultrasound experiment on the cranial CT image using a transducer array to obtain acoustic wave information. The acoustic wave information includes skull acoustic wave information, soft tissue acoustic wave information, and vascular acoustic wave information.

[0035] Design a transducer array to conduct ultrasound experiments on CT images of different parts of the brain, and acquire the data collected by the transducers. Simulation experiments can be conducted first to determine the specific design scheme of the transducers before actual experiments. Ultrasound experiments were performed on cranial CT images, including skull CT images, soft tissue CT images, and vascular CT images, to obtain skull acoustic wave information, soft tissue acoustic wave information, and vascular acoustic wave information.

[0036] Step 103: Invert the initial predicted signal wave and the acoustic wave information to obtain the inverted image. The initial predicted signal wave is obtained from an ultrasonic emission simulation experiment using a pure water model.

[0037] The inversion is performed based on the initial predicted signal wave and the acoustic wave information to obtain an inverted image. Specifically, this includes performing an inversion using a full waveform inversion algorithm based on the initial predicted signal wave and the acoustic wave information to obtain an inverted image.

[0038] The inversion is performed using a full waveform inversion algorithm based on the initial predicted signal wave and the acoustic wave information to obtain an inversion image. Specifically, this includes: performing forward modeling using the finite element method based on the initial predicted signal wave to obtain the predicted value; and performing iterative modeling using the steepest gradient method based on the predicted value and the acoustic wave information to obtain the inversion image.

[0039] like Figure 1 As shown, the inversion image is obtained by conducting an ultrasonic emission simulation experiment using a pure water model (denoted as s0) to obtain the initial predicted signal wave. An optimization algorithm is designed to iterate from the initial s0 to s to minimize the distance between the predicted value and the signal wave of the CT sound velocity model. The sound velocity s that minimizes this distance is used as the inversion result. That is, the real data obtained from the experiment is the sound wave information in step 102. Inversion steps are performed separately for the skull, soft tissue, and blood vessels, inverting the skull sound wave information, soft tissue sound wave information, and blood vessel sound wave information respectively to obtain inversion images for different parts. Since a deep learning network is used to construct the input image, the experiment needs to be repeated to generate the dataset.

[0040] Step 104: Register and fuse the inverted image with the cranial CT image as a reference to obtain a complete cranial inverted sound velocity map.

[0041] CT-assisted registration and fusion: The inverted images of the skull, brain tissue and blood vessels are registered and fused with CT as the reference. The registration is trained using a convolutional neural network, and the fusion is performed using the Laplacian pyramid method to fuse the inverted images to obtain a complete inverted sound velocity map of the brain.

[0042] Step 105: Using the complete inverted sound velocity map of the brain as input and the brain CT image as output, train the deep learning network to obtain the inversion model; the inversion model is used to invert the complete inverted sound velocity map of the brain.

[0043] Improving resolution using deep learning: Steps 103 and 104 are repeated extensively to construct the original dataset by registering and fusing the inverted cranial velocity maps and CT images. Then, using the CT images as the standard, a U-net neural network is constructed to improve the resolution and clarity of the inverted cranial velocity maps. In the actual inversion process, the input to the deep learning network is the complete inverted cranial velocity map, and the output is a high-resolution inverted cranial velocity map, where the high-resolution inverted cranial velocity map is an image with a quality close to that of the CT image.

[0044] This invention performs full-waveform inversion imaging of cranial tissues separately, and then constructs a deep learning network based on CT images to obtain high-quality inverted cranial images, thereby reducing the time required for inversion imaging. This invention has the following advantages: Reduced inversion time: The network learning time during inversion is less than that of traditional convergence methods such as steepest gradient descent. Improved resolution of different tissues: This invention performs inversion on different tissues separately, avoiding the situation where unclear images during the inversion process cause indistinguishability between tissues. Figure 3 This is a full waveform inversion image of brain tissue. The horizontal and vertical axes represent the image's location coordinates (a coordinate system established with the intersection of the central axes as the origin), and the unit is cm. Figure 3 In the diagram, (a) is the actual sound velocity map of brain tissue, in m / s; (b) is the actual brain tissue attenuation map, with the sound attenuation coefficient in dB / cm·MHz; (c) is the predicted sound velocity map of brain tissue obtained by the full waveform inversion method; and (d) is the predicted sound attenuation map of brain tissue obtained by the full waveform inversion method. Figure 3 It can be seen that the full waveform inversion algorithm can predict the sound velocity and attenuation distribution in brain tissue relatively well.

[0045] In one embodiment, a computer device is provided, which may be a database. The computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is connected to the system bus via the I / O interfaces. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device stores transactions to be processed. The I / O interfaces of the computer device are used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a data processing method.

[0046] In one embodiment, a computer device is also provided, including a memory and a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the above method embodiments.

[0047] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0048] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0049] It should be noted that the object information (including but not limited to object device information, object personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the object or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0050] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0051] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0052] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for full waveform inversion in two-dimensional cranial imaging, characterized in that, include: Acquire cranial CT images; the cranial CT images include skull CT images, soft tissue CT images, and vascular CT images; An ultrasound experiment was performed on the cranial CT image using a transducer array to obtain acoustic wave information; the acoustic wave information included skull acoustic wave information, soft tissue acoustic wave information, and vascular acoustic wave information. Inversion is performed based on the initial predicted signal wave and the acoustic wave information to obtain the inverted image, specifically including: Based on the initial predicted signal wave and the acoustic wave information, an inversion algorithm is used to perform an inversion to obtain an inverted image, specifically including: Based on the initial predicted value signal wave, the predicted value is obtained by performing forward modeling using the finite element method; Based on the predicted value and the acoustic information, the steepest gradient method is used for iteration to obtain the inverted image; The inverted image is registered and fused with the cranial CT image as a reference to obtain a complete cranial inverted sound velocity map. Using the complete inverted sound velocity map of the brain as input and the brain CT image as output, a deep learning network is trained to obtain an inversion model; the inversion model is used to invert the complete inverted sound velocity map of the brain.

2. The two-dimensional cranial imaging full waveform inversion method according to claim 1, characterized in that, The initial predicted signal wave was obtained from an ultrasonic emission simulation experiment using a pure water model.

3. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the steps of the method according to any one of claims 1-2.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-2.

5. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-2.

Images