CN117454692A - tDCS electric field simulation image generation method and device based on deep learning network - Google Patents
tDCS electric field simulation image generation method and device based on deep learning network Download PDFInfo
- Publication number
- CN117454692A CN117454692A CN202311369053.6A CN202311369053A CN117454692A CN 117454692 A CN117454692 A CN 117454692A CN 202311369053 A CN202311369053 A CN 202311369053A CN 117454692 A CN117454692 A CN 117454692A
- Authority
- CN
- China
- Prior art keywords
- tdcs
- magnetic resonance
- electric field
- image
- field simulation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 187
- 230000005684 electric field Effects 0.000 title claims abstract description 184
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000013135 deep learning Methods 0.000 title claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 76
- 238000004364 calculation method Methods 0.000 claims abstract description 55
- 238000012360 testing method Methods 0.000 claims description 29
- 238000010586 diagram Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000005457 optimization Methods 0.000 claims description 8
- 238000003709 image segmentation Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 abstract description 10
- 238000005516 engineering process Methods 0.000 abstract description 5
- 238000002595 magnetic resonance imaging Methods 0.000 description 28
- 210000003128 head Anatomy 0.000 description 26
- 210000001519 tissue Anatomy 0.000 description 12
- 230000000638 stimulation Effects 0.000 description 11
- 210000004556 brain Anatomy 0.000 description 6
- 210000003484 anatomy Anatomy 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 210000004761 scalp Anatomy 0.000 description 4
- 238000005094 computer simulation Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 238000011374 additional therapy Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 210000001175 cerebrospinal fluid Anatomy 0.000 description 2
- 230000019771 cognition Effects 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 210000004884 grey matter Anatomy 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004007 neuromodulation Effects 0.000 description 2
- 230000000704 physical effect Effects 0.000 description 2
- 208000020016 psychiatric disease Diseases 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 210000003625 skull Anatomy 0.000 description 2
- 210000004885 white matter Anatomy 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 235000002566 Capsicum Nutrition 0.000 description 1
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 1
- 241000282412 Homo Species 0.000 description 1
- 239000006002 Pepper Substances 0.000 description 1
- 235000016761 Piper aduncum Nutrition 0.000 description 1
- 235000017804 Piper guineense Nutrition 0.000 description 1
- 244000203593 Piper nigrum Species 0.000 description 1
- 235000008184 Piper nigrum Nutrition 0.000 description 1
- 208000028017 Psychotic disease Diseases 0.000 description 1
- FAPWRFPIFSIZLT-UHFFFAOYSA-M Sodium chloride Chemical compound [Na+].[Cl-] FAPWRFPIFSIZLT-UHFFFAOYSA-M 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 210000000988 bone and bone Anatomy 0.000 description 1
- 210000000845 cartilage Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000002872 contrast media Substances 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000004070 electrodeposition Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 210000003041 ligament Anatomy 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 230000007996 neuronal plasticity Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 239000011780 sodium chloride Substances 0.000 description 1
- 210000000278 spinal cord Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 210000001258 synovial membrane Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/10—Image enhancement or restoration using non-spatial domain filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a tDCS electric field simulation image generation method based on a deep learning network, which comprises the following steps: acquiring a plurality of sample magnetic resonance images; performing tDCS electric field simulation calculation on the sample magnetic resonance images by a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image; constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image; training a preset deep learning network model according to the training data set to obtain a target image generation model; and predicting the target magnetic resonance image through the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image. Therefore, the complicated electric field simulation calculation process before accurate positioning of the tDCS is reduced, and the electric field simulation efficiency is improved. Meanwhile, the technical requirement of clinicians on electric field simulation technology learning is omitted, and the feasibility of accurate positioning of tDCS and clinical application of personalized treatment is increased.
Description
Technical Field
The invention belongs to the technical field of medical image processing, and particularly relates to a tDCS electric field simulation image generation method, device and equipment based on a deep learning network and a storage medium.
Background
Transcranial direct current stimulation (transcranial direct current stimulation, tDCS) is a non-invasive neuromodulation technique that can promote brain neural plasticity and regulate human cognition and behavior. tDCS is well tolerated, can be used as an additional therapy for mental diseases, and is low cost, and has therefore become a focus of attention for clinicians as well as researchers. the tDCS can flexibly and reasonably select the number, the shape and the position of the electrodes according to the anatomical characteristics of the individual to realize specific anatomical positioning. The current electric field simulation calculation model has evolved from a simple concentric circle model to a high resolution anatomically accurate model based on individual anatomy magnetic resonance imaging (Magnetic Resonance Imaging, MRI) scanning. The high-resolution head model is generally constructed by taking image data of human brain in practice as a data source, and obtaining a final head model through processing and reconstructing the image data.
The existing tDCS electric field simulation modeling tools are varied, from finite element method models to independent GUI software designed for clinicians, and a plurality of software packages can be used for carrying out individual treatment design by loading individual MRI images. However, the current MRI image-based electric field simulation is relatively time-consuming, typically requiring at least 2 hours per subject, and is not efficient in implementing a personalized tDCS electric field simulation scheme clinically by conventional software. In addition, the adoption of the traditional software has a certain electric field simulation technical requirement for operators, and the accurate positioning of tDCS and the feasibility of individuation treatment in clinical application are poor.
Disclosure of Invention
In order to solve the problems in the related art, the invention provides a tDCS electric field simulation image generation method based on a deep learning network. The technical problems to be solved by the invention are realized by the following technical scheme:
in a first aspect, the present invention provides a tDCS electric field simulation image generation method based on a deep learning network, including: acquiring a plurality of sample magnetic resonance images; performing tDCS electric field simulation calculation on the sample magnetic resonance images by a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image; constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image; the training data set comprises a plurality of sample magnetic resonance images and tDCS electric field simulation images corresponding to each sample magnetic resonance image; training a preset deep learning network model according to the training data set to obtain a target image generation model; and predicting the target magnetic resonance image through a target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image.
In some possible embodiments, before performing tDCS electric field simulation calculation on the sample magnetic resonance images by a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image, the method further includes: optimizing the sample magnetic resonance image through a preset optimization algorithm to obtain an optimized sample magnetic resonance image; wherein the optimization algorithm includes median filtering and/or bias field correction.
In some possible embodiments, performing tDCS electric field simulation calculation on the sample magnetic resonance image by a preset tDCS electric field simulation calculation method to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image, including: performing tDCS electric field simulation calculation on each sample magnetic resonance image under different tDCS electrode configurations to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image; wherein the tDCS electrode configuration includes shape of the electrode, positional information of the electrode, and current magnitude information of the electrode.
In some possible embodiments, performing tDCS electric field simulation calculation on each sample magnetic resonance image under different tDCS electrode configurations to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image, including: image segmentation is carried out on each sample magnetic resonance image, and a plurality of single tissue structure images contained in each sample magnetic resonance image are obtained; constructing a head model of each sample magnetic resonance image according to the plurality of single tissue structure images; placing a tDCS electrode on the head model according to a preset tDCS electrode configuration; dividing the head model according to a plurality of single-organization structure images to obtain a finite element grid; finite element calculation is carried out on the finite element grid through a preset finite element solver and tDCS electrode configuration, and a finite element solving result is obtained; and visually displaying the finite element solving result to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
In some possible implementations, the training data set further includes a tDCS electrode configuration; training a preset deep learning network model according to a training data set to obtain a target image generation model, wherein the training data set comprises the following steps: training a preset deep learning network model according to the sample magnetic resonance image, the tDCS electrode configuration and the tDCS electric field simulation image corresponding to the sample magnetic resonance image under different tDCS electrode configurations to obtain a target image generation model.
In some possible embodiments, predicting the target magnetic resonance image by the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image includes: acquiring a target magnetic resonance image and a target tDCS electrode configuration; inputting the target magnetic resonance image and the target tDCS electrode configuration into a target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image under the target tDCS electrode configuration.
In some possible embodiments, before predicting the target magnetic resonance image by the target image generation model, obtaining a tDCS electric field simulation image of the target magnetic resonance image, the method further comprises: selecting a preset number of sample magnetic resonance images and corresponding tDCS electric field simulation images from the training data set as a test data set; inputting a sample magnetic resonance image in the test data set into a target image generation model to obtain a test tDCS electric field simulation image; evaluating the performance of the target image generation model according to the similarity between the test tDCS electric field simulation image and the corresponding tDCS electric field simulation image in the test data set; outputting the target image generation model when the performance of the target image generation model meets a preset threshold; and training the target image generation model again through the training data set when the performance of the target image generation model does not meet the preset threshold value, until the performance of the target image generation model meets the preset threshold value.
In a second aspect, the present invention provides a tDCS electric field simulation image generating apparatus based on a deep learning network, the apparatus comprising: a first acquisition module for acquiring a plurality of sample magnetic resonance images; the simulation calculation module is used for performing tDCS electric field simulation calculation on the sample magnetic resonance images through a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image; the data set construction module is used for constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image; the training data set comprises a plurality of sample magnetic resonance images and tDCS electric field simulation images corresponding to each sample magnetic resonance image; the training module is used for training a preset deep learning network model according to the training data set to obtain a target image generation model; and the prediction module is used for predicting the target magnetic resonance image through the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image.
In some possible embodiments, the apparatus further comprises: the optimizing module is used for optimizing the sample magnetic resonance image through a preset optimizing algorithm to obtain an optimized sample magnetic resonance image; wherein the optimization algorithm includes median filtering and/or bias field correction.
In some possible embodiments, the first obtaining module is further configured to perform tDCS electric field simulation calculation on each sample magnetic resonance image under different tDCS electrode configurations, to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image; wherein the tDCS electrode configuration includes shape of the electrode, positional information of the electrode, and current magnitude information of the electrode.
In some possible embodiments, the first acquiring module is further configured to perform image segmentation on each sample magnetic resonance image to obtain a plurality of single tissue structure images included in each sample magnetic resonance image; constructing a head model of each sample magnetic resonance image according to the plurality of single tissue structure images; placing a tDCS electrode on the head model according to a preset tDCS electrode configuration; dividing the head model according to a plurality of single-organization structure images to obtain a finite element grid; finite element calculation is carried out on the finite element grid through a preset finite element solver and tDCS electrode configuration, and a finite element solving result is obtained; and visually displaying the finite element solving result to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
In some possible implementations, the training data set further includes a tDCS electrode configuration; the training module is also used for training a preset deep learning network model according to the sample magnetic resonance image, the tDCS electrode configuration and the tDCS electric field simulation image corresponding to the sample magnetic resonance image under different tDCS electrode configurations to obtain a target image generation model.
In some possible implementations, the prediction module is further configured to acquire a target magnetic resonance image and a target tDCS electrode configuration; inputting the target magnetic resonance image and the target tDCS electrode configuration into a target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image under the target tDCS electrode configuration.
In some possible embodiments, the apparatus further comprises: the test module is used for selecting a preset number of sample magnetic resonance images and corresponding tDCS electric field simulation images from the training data set to serve as a test data set; inputting a sample magnetic resonance image in the test data set into a target image generation model to obtain a test tDCS electric field simulation image; evaluating the performance of the target image generation model according to the similarity between the test tDCS electric field simulation image and the corresponding tDCS electric field simulation image in the test data set; outputting the target image generation model when the performance of the target image generation model meets a preset threshold; and training the target image generation model again through the training data set when the performance of the target image generation model does not meet the preset threshold value, until the performance of the target image generation model meets the preset threshold value.
In a third aspect, the present invention provides a tDCS electric field simulation image generating apparatus based on a deep learning network, including: a memory storing computer executable instructions; a processor, coupled to the memory, for executing computer-executable instructions to perform a method as in the first aspect of the invention and possible embodiments thereof.
In a fourth aspect, the present invention provides a computer storage medium having stored thereon computer executable instructions which, when executed by a processor, enable the implementation of a method as in the first aspect of the present invention and possible embodiments thereof.
Compared with the prior art, the technical scheme provided by the embodiment of the invention has the beneficial effects that:
in the invention, a tDCS electric field simulation image corresponding to each sample magnetic resonance image is obtained by performing tDCS electric field simulation calculation on a plurality of sample magnetic resonance images, so that a training data set is constructed to train a preset deep learning network model, a target image generation model is obtained, and the target magnetic resonance image is predicted by using the target image generation model, so that the tDCS electric field simulation image of the target magnetic resonance image is obtained. Therefore, the complicated electric field simulation calculation process before accurate positioning of the tDCS is reduced, the electric field simulation time cost is saved, and the electric field simulation efficiency is improved. Meanwhile, a target image generation model is constructed through deep learning, so that the technical requirement of a clinician on electric field simulation technology learning is omitted, the feasibility of accurate positioning of tDCS and clinical application of personalized treatment is increased, and the practical significance of a personalized tDCS treatment scheme based on an individual anatomical structure is improved.
The present invention will be described in further detail with reference to the accompanying drawings and examples.
Drawings
FIG. 1 is a schematic flow chart of a method for generating a tDCS electric field simulation image based on a deep learning network according to an embodiment of the present invention;
FIG. 2a is a schematic diagram of a conventional electrode configuration of a tDCS;
FIG. 2b is a schematic diagram of a model of tDCS under high precision HD-tDCS
FIG. 3 is a schematic diagram of an electric field simulation flow in an embodiment of the present invention;
FIG. 4 is a schematic diagram of a tDCS electric field simulation image according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of a tDCS electric field simulation image generating device based on a deep learning network in an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to specific examples, but embodiments of the present invention are not limited thereto.
In the description of the present invention, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Further, one skilled in the art can engage and combine the different embodiments or examples described in this specification.
Although the invention is described herein in connection with various embodiments, other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the "a" or "an" does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.
In order to illustrate the technical scheme of the invention, the following description is made by specific examples.
Transcranial direct current stimulation (transcranial direct current stimulation, tDCS) is a non-invasive neuromodulation technique that has evolved as an effective tool for inducing neuroplasticity and regulating cognition and behavior in humans. tDCS is a weak direct current stimulus, direct current is conducted through electrode pads, which are typically made of conductive rubber or metal electrodes, and then wrapped with a sponge sheet soaked in saline. In addition, the direct current can be conducted by directly contacting the electrode with the skin and conducting the current through the conductive gel as a conducting medium. The stimulation current is usually 1-2mA, the stimulation time is 20min, and the electrode size is 3.5-100cm2. There is also a 4 x 1 high precision tDCS (HD-tDCS) with more focusing, the center electrode being the anode, four cathodes arranged around the anode, the anode being the inflow electrode and the cathode being the outflow electrode. The conventional tDCS applies weak direct current to the scalp through a rectangular or circular sponge electrode sheet, and the high-precision tDCS stimulates the scalp using a small electrode array.
There is growing evidence that the delivery of electrical current to specific brain regions can promote plastic changes in the brain, and that this non-invasive stimulation technique is becoming a reliable means of effectively treating psychotic symptoms. However, the stimulation is performed within safe and acceptable parameters at the time of use. Because of its good tolerability, tDCS can be used as an additional therapy for mental diseases and is low cost to continue to use, it has become a focus of attention for clinicians as well as researchers. the tDCS can flexibly and reasonably select the number, the shape and the position of the electrodes according to the anatomical characteristics of the individual to realize specific anatomical positioning. Electric field simulation computational models have evolved from simple concentric circular models to high resolution anatomically accurate models based on individual anatomy Magnetic Resonance Imaging (MRI) scans. At present, the construction of a high-resolution head model generally takes image data of human brain in practice as a data source, and a final head model is obtained through processing and reconstructing the image data.
With the rapid development of deep learning in various fields, many scholars are led to study the application of deep learning in the field of medical images, and prediction of medical images is an important study direction of deep learning in the medical field. By analyzing and predicting the medical image by using the deep learning model, the medical image diagnosis system can help doctors to carry out tasks such as disease diagnosis, treatment planning and the like.
The existing tDCS electric field simulation modeling tools are varied, from finite element method models to independent GUI software designed for clinicians, and a plurality of software packages can be used for carrying out individual treatment design by loading individual MRI images. However, the current MRI image-based electric field simulation is relatively time-consuming, typically requiring at least 2 hours per subject, and is not efficient in implementing a personalized tDCS electric field simulation scheme clinically by conventional software. In addition, the adoption of traditional software has certain electric field simulation technical requirements for operators, medical image prediction based on deep learning needs a large-scale and high-quality medical image data set, and a professional doctor is required to accurately mark images, so that accurate positioning of tDCS and feasibility of personalized treatment in clinical application are poor.
In view of the above, the embodiments of the present invention provide a method for generating a tDCS electric field simulation image based on a deep learning network, so as to solve the above problems.
Referring to fig. 1, fig. 1 is a schematic flow chart of a tDCS electric field simulation image generating method based on a deep learning network according to an embodiment of the present invention, and the following description will refer to the steps shown in fig. 1.
S101, acquiring a plurality of sample magnetic resonance images;
wherein, individual anatomy magnetic resonance imaging (Magnetic Resonance Imaging, MRI) is a safe and reliable high-tech examination device, has no X-ray radiation and no harm to human body. As a high-end core technology of medical imaging, MRI has been in clinical application history for nearly 30 years, technology has been rapidly developed, hardware platforms and software technologies are continuously updated, and clinical application fields are gradually expanded. MRI images are very fine, sharp, realistic. MRI examination has the advantages of no X-ray radiation, no contrast agent for clearly displaying heart, blood vessel and internal cavity, accurate positioning of any orientation tomographic scan, etc. The MRI clinical indication is extensive, is the first choice examination method of the lesions of the craniocerebral, spinal cord, bone and joint cartilage, synovium, ligament, etc.
It will be appreciated that the conventional MRI images are typically DICOM files in two-dimensional format, i.e., the plurality of sample magnetic resonance images in embodiments of the present invention are DICOM files. In order to facilitate subsequent processing, in the embodiment of the present invention, the obtained DICOM file may be converted into a three-dimensional format NIfTI file.
Based on this, in some embodiments, before performing the above step S102, the above method may further perform the following step S11:
s11, optimizing a sample magnetic resonance image through a preset optimization algorithm to obtain an optimized sample magnetic resonance image;
wherein the optimization algorithm includes median filtering and/or bias field correction.
It will be appreciated that during the acquisition of MRI images, the acquired MRI images may be subject to noise and distortion due to random interference and the like that is typically present in the imaging device. By using median filtering on the MRI image, salt and pepper noise in the MRI image can be removed to enhance the image effect, and edges and details of the image can be well preserved. In addition, gray scale non-uniformity may be caused when using magnetic resonance scanning, and the imaging quality of the final structural image can be improved by correcting the MRI image by the bias field.
S102, performing tDCS electric field simulation calculation on a sample magnetic resonance image by a preset tDCS electric field simulation calculation method to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image;
in some embodiments, the step S102 may further include the following step S22:
s22, performing tDCS electric field simulation calculation on each sample magnetic resonance image under different tDCS electrode configurations to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
Wherein the tDCS electrode configuration includes shape of the electrode, positional information of the electrode, and current magnitude information of the electrode.
It can be understood that the application of the electrical stimulation to the head through the tDCS is achieved by setting electrode plates on the head, and the electrical stimulation has different effects due to the shape, the setting position, the magnitude of the applied current and other factors of different electrode plates, so that the final imaging result is affected. Therefore, under different electrode configurations, different tDCS electric field simulation images can be obtained by carrying out the tDCS electric field simulation calculation on the same sample magnetic resonance image.
Exemplary, fig. 2 is a schematic diagram of a tDCS model under different electrode configurations in an embodiment of the present invention, where fig. 2a is a schematic diagram of a tDCS model under a conventional electrode configuration, where the electrode configuration is based on a standard lead system, and each electrode position is a cathode-anode electrode pair formed by any two electrodes in the conventional standard lead system. Fig. 2b is a schematic diagram of a model of tDCS under high precision HD-tDCS, where the anode a position is any one electrode in a standard lead system, and four cathodes b are around the anode. The current levels in fig. 2a and 2b may be preset current levels, and the current levels in fig. 2a and 2b may be the same or different.
In some embodiments, the step S102 may further include the following steps:
s1021, performing image segmentation on each sample magnetic resonance image to obtain a plurality of single tissue structure images contained in each sample magnetic resonance image;
s1022, constructing a head model of each sample magnetic resonance image according to the plurality of single tissue structure images;
s1023, placing a tDCS electrode on the head model according to a preset tDCS electrode configuration;
s1024, dividing the head model according to a plurality of single-organization structure images to obtain finite element grids;
s1025, carrying out finite element calculation on the finite element grid through a preset finite element solver and tDCS electrode configuration to obtain a finite element solving result;
and S1026, visually displaying the finite element solving result to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
In some embodiments, the plurality of single tissue structure images may include gray matter, white matter, cerebrospinal fluid, scalp, skull, and air. From these six partial single organization structures, the construction of the head model can be performed.
Fig. 3 is a schematic structural diagram of an electric field simulation flow in an embodiment of the present invention, referring to fig. 3, after an original MRI image is subjected to image segmentation, a single tissue structure including six parts including gray matter, white matter, cerebrospinal fluid, scalp, skull and air is obtained, then a head model is constructed according to the six single tissue structures, an electrode is placed on the head model, then finite element calculation is performed, and simulation is performed according to the finite element calculation result, so as to obtain a final simulation result, that is, a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
In some embodiments, the physical properties of the electrode may include the shape and size of the electrode, such as the electrode shown in fig. 2a or 2b, when the electrode is placed on the head model, which may be by accurately modeling the physical properties of the electrode. After the electrodes are accurately modeled, they are accurately placed on the head model. The finite element mesh generation process mainly generates an accurate finite element mesh from a single tissue structure and an electrode mask, and meanwhile, the resolution of an anatomical image is required. During this grid generation, each mask may be divided into small, successive elements, so that a numerical calculation of the current may be achieved. The finite element solving result is visually displayed, the finite element solving result can be displayed in an image or numerical mode, and further the change condition of related variables of a research object and the relation among various physical quantities can be analyzed in an intuitive mode.
It will be appreciated that, through the above step S102, an electric field distribution result may be obtained, for example, an MRI structural image of a patient is obtained, and the structural image is first subjected to image preprocessing, and then a head model is built, and electrodes are placed on the head model, so as to simulate tDCS treatment on a real patient' S head. Referring to fig. 4, fig. 4 is a schematic diagram of a tDCS electric field simulation image according to an embodiment of the present invention, where the electrode configuration may be set as follows: the anode is placed at F3, the cathode is placed at F4, or the anode F3, the cathode is wrapped around F3, and a stimulus current of 2mA is applied. After finite element calculation, a schematic diagram of a tDCS electric field simulation image shown in fig. 4 can be obtained.
It should be noted that different electrode configurations may generate different electrical stimulation effects on the brain, so the values of the specific electrode configurations may be set based on the needs in the practical application process, which is not particularly limited in the embodiment of the present invention.
S103, constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image;
the training data set comprises a plurality of sample magnetic resonance images and tDCS electric field simulation images corresponding to each sample magnetic resonance image.
S104, training a preset deep learning network model according to the training data set to obtain a target image generation model;
it can be understood that, after performing tDCS electric field simulation calculation on each sample magnetic resonance image in step S102, a tDCS electric field simulation image corresponding to each sample magnetic resonance image can be obtained correspondingly. In this way, each sample magnetic resonance image and the tDCS electric field simulation image corresponding to each sample magnetic resonance image are used as training data, so that a training data set is constructed to train a preset model, and a target image generation model for generating the tDCS electric field simulation image by converting the magnetic resonance image can be obtained.
In some embodiments, since the tDCS electric field simulation image obtained by performing tDCS electric field simulation calculation on the sample magnetic resonance image in step S102 is different under different electrode configurations, the influence of the electrode configurations may also be considered in the process of constructing the training data set in step S103. Thus, in an embodiment of the present invention, the training data set may further include a tDCS electrode configuration.
Based on this, the above step S103 may further include the following step S1031:
s1031, training a preset deep learning network model according to the sample magnetic resonance image, the tDCS electrode configuration and the tDCS electric field simulation image corresponding to the sample magnetic resonance image under different tDCS electrode configurations to obtain a target image generation model.
It can be appreciated that under different electrode configurations, the tDCS electric field simulation image obtained by performing the tDCS electric field simulation calculation on the sample magnetic resonance image in step S102 is different. Then, the electrode configuration is also used as training data to train the model, and the obtained target network model can convert the MRI image into the tDCS electric field simulation image corresponding to the MRI image under the corresponding electrode configuration according to the electrode configuration in the process of generating the tDCS electric field simulation image according to the original MRI image. In this way, the performance of the target image generation model can be improved.
In some embodiments, the architecture of the pre-set deep learning network model may be a network based on a convolutional network Unet improvement for biomedical image segmentation. The improved network architecture may employ an attention mechanism (Attention Mechanism) by which the deep learning model may be focused more on the most relevant regions or features in the medical image, improving the predictive performance of the target image generation model. Each layer of convolution of the original Unet network can be added with a layer of Residual convolution layer (Residual Block) connected with Residual, so that training and convergence effects of the network can be improved, and information fusion is carried out to further improve prediction accuracy of a target image generation model.
In some embodiments, model parameters may be optimized by an adaptive moment estimation (Adaptive Moment Estimation, adam) algorithm during training of a pre-set deep-learning network model to provide better convergence performance and generalization capability. The loss function may be a cross entropy loss function (Cross Entropy Loss, CE), and may measure the difference between the simulation result output by the model in the training process and the real simulation result label, and set the learning rate to make the model converge rapidly and finally improve the performance of the target image generation model.
It will be appreciated that after training the model with the training dataset to obtain the target image generation model, the original MRI image can be directly converted into a tDCS electric field simulation image by the target image generation model. However, in the case where the performance of the target image generation model is low, there may be a case where the tDCS electric field simulation image generated directly through the target image generation model does not meet the accuracy requirement in the actual application process. At this time, the training data set may be used to iteratively train the obtained target image generation model, so as to improve the performance of the target image generation model.
Based on this, in some embodiments, after the target image generation model is obtained through the step S104, the method may further include the steps of:
s1041, selecting a preset number of sample magnetic resonance images and corresponding tDCS electric field simulation images from a training data set as a test data set;
s1042, inputting a sample magnetic resonance image in the test data set into a target image generation model to obtain a test tDCS electric field simulation image;
s1043, evaluating the performance of the target image generation model according to the similarity between the test tDCS electric field simulation image and the corresponding tDCS electric field simulation image in the test data set;
S1044, outputting a target image generation model when the performance of the target image generation model meets a preset threshold;
s1045, when the performance of the target image generation model does not meet the preset threshold, training the target image generation model again through the training data set until the performance of the target image generation model meets the preset threshold.
It can be appreciated that the tDCS electric field simulation image in the training data set is obtained by the tDCS electric field simulation calculation method. And selecting a preset number of sample magnetic resonance images and corresponding tDCS electric field simulation images from the training data set as a test data set, and if the similarity between the test tDCS electric field simulation images obtained through the target image generation model and the corresponding tDCS electric field simulation images in the training data set meets the preset threshold, indicating that the tDCS electric field simulation images obtained through the target image generation model and the tDCS electric field simulation images obtained through the tDCS electric field simulation calculation method have similar effects. That is, the target image generation model can replace the calculation of the tDCS electric field simulation, so that the complicated electric field simulation calculation process before the accurate positioning of the tDCS is omitted, and the electric field simulation time cost is saved.
S105, predicting the target magnetic resonance image through the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image.
In some embodiments, the step S105 may include the following steps:
s1051, acquiring a target magnetic resonance image and a target tDCS electrode configuration;
s1052, inputting the target magnetic resonance image and the target tDCS electrode configuration into a target image generation model, and obtaining a tDCS electric field simulation image of the target magnetic resonance image under the target tDCS electrode configuration.
The target magnetic resonance image is an MRI image in which tDCS electric field simulation is required, and the target image generation model is the target image generation model obtained in steps S101 to S104. Thus, for any target magnetic resonance image, only the required target tDCS electrode configuration needs to be acquired, and the target magnetic resonance image can be converted into a tDCS electric field simulation image under the target tDCS electrode configuration through the target image generation model.
In the embodiment of the invention, the tDCS electric field simulation image corresponding to each sample magnetic resonance image is obtained by performing tDCS electric field simulation calculation on a plurality of sample magnetic resonance images, so that a training data set is constructed to train a preset deep learning network model, a target image generation model is obtained, and the target magnetic resonance image is predicted by using the target image generation model, so that the tDCS electric field simulation image of the target magnetic resonance image is obtained. Therefore, the complicated electric field simulation calculation process before accurate positioning of the tDCS is reduced, the electric field simulation time cost is saved, and the electric field simulation efficiency is improved. Meanwhile, a target image generation model is built through deep learning, so that the technical requirement of a clinician on electric field simulation technology learning is omitted, the accurate positioning of tDCS and the feasibility of clinical application of personalized treatment are increased, and the practical significance of a personalized tDCS treatment scheme based on an individual anatomical structure is improved.
Fig. 5 is a schematic structural diagram of a tDCS electric field simulation image generating apparatus based on a deep learning network in an embodiment of the present invention, and referring to fig. 5, the apparatus 500 may include:
a first acquisition module 501 for acquiring a plurality of sample magnetic resonance images;
the simulation calculation module 502 is configured to perform tDCS electric field simulation calculation on the sample magnetic resonance images by using a preset tDCS electric field simulation calculation method, so as to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image;
a data set construction module 503 for constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image; the training data set comprises a plurality of sample magnetic resonance images and tDCS electric field simulation images corresponding to each sample magnetic resonance image;
the training module 504 is configured to train a preset deep learning network model according to the training data set, so as to obtain a target image generation model;
the prediction module 505 is configured to predict the target magnetic resonance image through the target image generation model, and obtain a tDCS electric field simulation image of the target magnetic resonance image.
In some possible embodiments, the apparatus 500 further comprises: the optimizing module is used for optimizing the sample magnetic resonance image through a preset optimizing algorithm to obtain an optimized sample magnetic resonance image; wherein the optimization algorithm includes median filtering and/or bias field correction.
In some possible embodiments, the first obtaining module 501 is further configured to perform tDCS electric field simulation calculation on each sample magnetic resonance image under different tDCS electrode configurations, to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image; wherein the tDCS electrode configuration includes shape of the electrode, positional information of the electrode, and current magnitude information of the electrode.
In some possible embodiments, the first obtaining module 501 is further configured to perform image segmentation on each sample magnetic resonance image to obtain a plurality of single tissue structure images contained in each sample magnetic resonance image; constructing a head model of each sample magnetic resonance image according to the plurality of single tissue structure images; placing a tDCS electrode on the head model according to a preset tDCS electrode configuration; dividing the head model according to a plurality of single-organization structure images to obtain a finite element grid; finite element calculation is carried out on the finite element grid through a preset finite element solver and tDCS electrode configuration, and a finite element solving result is obtained; and visually displaying the finite element solving result to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
In some possible implementations, the training data set further includes a tDCS electrode configuration; the training module 504 is further configured to train a preset deep learning network model according to the sample magnetic resonance image, the tDCS electrode configuration, and tDCS electric field simulation images corresponding to the sample magnetic resonance image under different tDCS electrode configurations, so as to obtain a target image generation model.
In some possible implementations, the prediction module 505 is further configured to acquire a target magnetic resonance image and a target tDCS electrode configuration; inputting the target magnetic resonance image and the target tDCS electrode configuration into a target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image under the target tDCS electrode configuration.
In some possible embodiments, the apparatus 500 further comprises: the test module is used for selecting a preset number of sample magnetic resonance images and corresponding tDCS electric field simulation images from the training data set to serve as a test data set; inputting a sample magnetic resonance image in the test data set into a target image generation model to obtain a test tDCS electric field simulation image; evaluating the performance of the target image generation model according to the similarity between the test tDCS electric field simulation image and the corresponding tDCS electric field simulation image in the test data set; outputting the target image generation model when the performance of the target image generation model meets a preset threshold; and training the target image generation model again through the training data set when the performance of the target image generation model does not meet the preset threshold value, until the performance of the target image generation model meets the preset threshold value.
Based on the same inventive concept, the embodiments of the present invention provide an electronic device, which may be consistent with the tDCS electric field simulation image generation method based on the deep learning network described in one or more of the embodiments. Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and referring to fig. 6, an electronic device 600 may be a general-purpose computer hardware including a processor 601 and a memory 602
Based on the same inventive concept, the invention provides a computer storage medium, which stores computer executable instructions that, when executed by a processor, can implement the tDCS electric field simulation image generation method based on the deep learning network according to one or more embodiments.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and it is not intended that the invention be limited to the specific embodiments described. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.
Claims (10)
1. The method for generating the tDCS electric field simulation image based on the deep learning network is characterized by comprising the following steps of:
acquiring a plurality of sample magnetic resonance images;
performing tDCS electric field simulation calculation on the sample magnetic resonance images by a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image;
constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image; the training data set comprises a plurality of sample magnetic resonance images and the tDCS electric field simulation images corresponding to each sample magnetic resonance image;
training a preset deep learning network model according to the training data set to obtain a target image generation model;
and predicting the target magnetic resonance image through the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image.
2. The simulated image generation method of claim 1, wherein prior to performing tDCS electric field simulation calculation on the sample magnetic resonance images by a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulated images corresponding to each of the sample magnetic resonance images, the method further comprises:
Optimizing the sample magnetic resonance image through a preset optimization algorithm to obtain an optimized sample magnetic resonance image; wherein the optimization algorithm comprises median filtering and/or bias field correction.
3. The method for generating a simulation image according to claim 1, wherein the performing tDCS electric field simulation calculation on the sample magnetic resonance image by a preset tDCS electric field simulation calculation method to obtain a tDCS electric field simulation image corresponding to each of the sample magnetic resonance images includes:
performing tDCS electric field simulation calculation on each sample magnetic resonance image under different tDCS electrode configurations to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image; wherein the tDCS electrode configuration includes shape of the electrode, positional information of the electrode, and current magnitude information of the electrode.
4. A simulated image generation method as claimed in claim 3, wherein said performing tDCS electric field simulation calculations on each of said sample magnetic resonance images under different tDCS electrode configurations to obtain a tDCS electric field simulated image corresponding to each of said sample magnetic resonance images comprises:
image segmentation is carried out on each sample magnetic resonance image, and a plurality of single tissue structure images contained in each sample magnetic resonance image are obtained;
Constructing a head model of each of the sample magnetic resonance images from the plurality of single tissue structure images;
placing a tDCS electrode on the head model according to a preset tDCS electrode configuration;
dividing the head model according to the plurality of single-organization chart diagrams to obtain a finite element grid;
performing finite element calculation on the finite element grid through a preset finite element solver and the tDCS electrode configuration to obtain a finite element solving result;
and visually displaying the finite element solving result to obtain a tDCS electric field simulation image corresponding to each sample magnetic resonance image.
5. A simulated image generation method as claimed in claim 3 wherein said training dataset further comprises said tDCS electrode configuration;
training a preset deep learning network model according to the training data set to obtain a target image generation model, wherein the training comprises the following steps:
training the preset deep learning network model according to the sample magnetic resonance image, the tDCS electrode configuration and tDCS electric field simulation images corresponding to the sample magnetic resonance image under different tDCS electrode configurations to obtain the target image generation model.
6. The image generation method according to claim 5, wherein predicting a target magnetic resonance image by the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image, comprises:
acquiring the target magnetic resonance image and target tDCS electrode configuration;
inputting the target magnetic resonance image and the target tDCS electrode configuration into the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image under the target tDCS electrode configuration.
7. The image generation method according to claim 6, characterized in that before predicting a target magnetic resonance image by the target image generation model, obtaining a tDCS electric field simulation image of the target magnetic resonance image, the method further comprises:
selecting a preset number of sample magnetic resonance images and corresponding tDCS electric field simulation images from the training data set as a test data set;
inputting a sample magnetic resonance image in the test data set into the target image generation model to obtain a test tDCS electric field simulation image;
evaluating the performance of the target image generation model according to the similarity between the test tDCS electric field simulation image and the corresponding tDCS electric field simulation image in the test data set;
Outputting the target image generation model when the performance of the target image generation model meets a preset threshold;
and training the target image generation model again through the training data set when the performance of the target image generation model does not meet the preset threshold value until the performance of the target image generation model meets the preset threshold value.
8. A tDCS electric field simulation image generation device based on a deep learning network, comprising:
a first acquisition module for acquiring a plurality of sample magnetic resonance images;
the simulation calculation module is used for performing tDCS electric field simulation calculation on the sample magnetic resonance images through a preset tDCS electric field simulation calculation method to obtain tDCS electric field simulation images corresponding to each sample magnetic resonance image;
the data set construction module is used for constructing a training data set according to the sample magnetic resonance image and the tDCS electric field simulation image; the training data set comprises a plurality of sample magnetic resonance images and the tDCS electric field simulation images corresponding to each sample magnetic resonance image;
the training module is used for training a preset deep learning network model according to the training data set to obtain a target image generation model;
And the prediction module is used for predicting the target magnetic resonance image through the target image generation model to obtain a tDCS electric field simulation image of the target magnetic resonance image.
9. An electronic device, comprising:
a memory storing computer executable instructions;
a processor, coupled to the memory, for implementing the method of any one of claims 1 to 7 by executing the computer-executable instructions.
10. A computer storage medium storing computer executable instructions which, when executed by a processor, enable the method of any one of claims 1 to 7 to be carried out.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311369053.6A CN117454692A (en) | 2023-10-20 | 2023-10-20 | tDCS electric field simulation image generation method and device based on deep learning network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311369053.6A CN117454692A (en) | 2023-10-20 | 2023-10-20 | tDCS electric field simulation image generation method and device based on deep learning network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117454692A true CN117454692A (en) | 2024-01-26 |
Family
ID=89588442
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311369053.6A Pending CN117454692A (en) | 2023-10-20 | 2023-10-20 | tDCS electric field simulation image generation method and device based on deep learning network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117454692A (en) |
-
2023
- 2023-10-20 CN CN202311369053.6A patent/CN117454692A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7405818B2 (en) | Treat the patient with a TT field with optimized electrode positions using a deformable template | |
| JP7383679B2 (en) | TTFIELD treatment with optimized electrode position on the head based on MRI conductivity measurements | |
| JP7271000B2 (en) | Patch guide method and program | |
| JP7263324B2 (en) | Method and program for generating 3D brain map | |
| CN110337312B (en) | Selection and placement of optimal deep brain stimulation electrodes based on stimulation field modeling | |
| CN111311703B (en) | Electrical impedance tomography image reconstruction method based on deep learning | |
| CN105263405A (en) | Multi-parameter physiological mapping | |
| EP4003146A1 (en) | Systems and methods for determining a fluid and tissue volume estimations using electrical property tomography | |
| CN114463493A (en) | Transcranial magnetic stimulation electric field rapid imaging method and model based on coding and decoding structure | |
| Gomez et al. | Analysis of fiber strain in the human tongue during speech | |
| CN117454692A (en) | tDCS electric field simulation image generation method and device based on deep learning network | |
| CN114596225A (en) | Motion artifact simulation method and system | |
| Frangi et al. | Special Issue on Multiscale Modeling and Analysis in Computational Biology and Medicine—Part-1 | |
| KR102373760B1 (en) | System and method for executing brain stimulus simulation according to pre-set guide system using external server based on anonymized data | |
| US11291832B2 (en) | Patient-specific local field potential model | |
| CN117934726B (en) | Three-dimensional visualization method, apparatus and system, and readable storage medium | |
| WO2021206157A1 (en) | Medical information processing device and medical information processing method | |
| JP2024506509A (en) | Methods, systems and apparatus for medical image enhancement that optimize transducer array placement | |
| Taberna | Advanced head modelling methods for the study of neural connectivity | |
| KR20230008939A (en) | System and method for executing brain stimulus simulation according to pre-set guide system using external server based on anonymized data | |
| JP2021166695A (en) | Medical information processing device and medical information processing method | |
| CN116894356A (en) | Human brain personalized time interference stimulation simulation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |