37 results about "Neural imaging" patented technology

Apparatus and methods for non-invasive electrical stimulation of the brain through skin surface stimulation of the peripheral nervous system as a treatment for movement disorders. Skin surface electrodes are positioned at predetermined peripheral surface stimulation sites on the skin surface using a variety of neural imaging techniques. A pulsatile electrical current is generated at the stimulation sites through a variety of standard electrical stimulation devices. Stimulation of the peripheral surface stimulation sites translates to electrical stimulation of a specific area of the brain.

The invention relates to an intracranial puncture locating method based on mixed reality and belongs to the field of neural imaging navigation. The method includes the steps of firstly, collecting CTscanning data of a patient with a marker, and creating three-dimensional models of the head and intracranial lesions of the patient according to the CT data; determining puncture targets, a puncture direction and puncture depth on the basis of the three-dimensional models, and creating an intracranial puncture path; finally, through rigid registration, overlapping the three-dimensional model of the head and the puncture path on a surgical part of the patient through a mixed reality technology, so that a surgeon can conduct puncture surgery under the guidance of the virtual three-dimensional models and correct puncture operation in real time according to the puncture path. Since the virtual three-dimensional modules are totally matched with a real scene, the virtual puncture patch is highlymatched with an actual situation, and therefore by conducting the puncture operation along the virtual puncture path, deviations or faults existing in traditional empirical operation are effectivelyavoided so that the purpose of improving the intracranial puncture precision and puncture efficiency can be achieved.

An encephalic electrode individualization locating method based on multimode medical image data fusion relates to the fields of medical instruments and medical image processing and neuroimaging. By means of multimode medical image data including magnetic resonance images, X-ray images and intranperative optical photos, the space position relation of an encephalic electrode and a brain tissue structure is comprehensively measured, a relation between the position of the encephalic electrode and a brain anatomical structure is established, and accurate and fast individualization locating of the encephalic electrode is achieved. The encephalic electrode individualization locating method based on multimode medical image data fusion can be used in the fields of cognitive neuroscience basic research, clinic neurosciences and the like, provide accurate brain space position information for an encephalic electroencephalogram signal, and enhance application value in the aspect of electrophysiological basis research.

The invention provides an analysis method for infant brain medical computer scanning images and a realization system. The analysis method is to process the brain medical computer scanning images into images with prominent fractal characteristics, and then perform mathematic quantitative analysis on the images by a chaos fractal analysis method. The method is realized under Matlab R2007 through programming, utilizes chaotic neural network and fractal principles, performs serialized division, identification and fractal mode analysis on the infant brain medical computer scanning images to obtain fractal dimension quantified reference values and quantified reference values of multifractal spectrum width of normal infant brain medical computer scanning images in different age brackets, realizes clinical prediction on sick infants with intelligent disability and brain paralysis which have no typical neural image manifest characteristics by taking the reference values as a comparison standard, thoroughly changes the conventional analysis mode for the infant brain medical computer scanning images, and has high accuracy and strong repeatability compared with scoring results of the clinically widely used Gesell scale.

Apparatus and methods for non-invasive electrical stimulation of the brain through skin surface stimulation of the peripheral nervous system as a treatment for movement disorders. Skin surface electrodes are positioned at predetermined peripheral surface stimulation sites on the skin surface using a variety of neural imaging techniques. A pulsatile electrical current is generated at the stimulation sites through a variety of standard electrical stimulation devices. Stimulation of the peripheral surface stimulation sites translates to electrical stimulation of a specific area of the brain.

The invention belongs to the technical field of neuroimaging data analysis and discloses a principal component analysis-based key encephalic region measurement method. The method comprises collecting and preprocessing fMRI (functional magnetic resonance imaging) data of M persons in a resting state; structuring an encephalic network under Q sparseness; calculating the weighted node degree, the node efficiency and the betweenness centrality of the encephalic network at every sparseness; through principal component analysis, performing comprehensive dimension reduction on the obtained node degrees, the node efficiency and the betweenness centralities to obtain a new index, namely, the centrality score H of every node of every person at every sparseness; integrating the Q sparseness and the centrality scores of the M persons, acquiring the centrality score of every node, and according to the centrality score of every node, determining key encephalic regions. The principal component analysis-based key encephalic region measurement method is applied to measuring the importance of encephalic regions, overcomes the problem of active selection of determining the sparseness of functional encephalic networks, and through combination of various classical encephalic region importance measurement methods, achieves more objective and more accurate key encephalic region measurement.

The invention discloses a flexible implantable neural photoelectrode. The flexible implantable neural photoelectrode comprises a recording electrode layer, a metal interconnection layer and an opticaldevice layer which are sequentially arranged, wherein a plurality of electrode sites are arranged on the recording electrode layer, the metal interconnection layer is used for connecting the electrode sites with the rear end of a neural imaging system, and the optical device layer is used for transmitting laser from the rear end of the neural imaging system to the flexible implantable neural photoelectrode and emitting the laser. The invention further discloses a method for preparing the flexible implantable neural photoelectrode. The flexible implantable neural photoelectrode provided by theinvention is high in integration level, and multi-channel and high-density signal recording can be realized; the size of the part implanted into the brain is small, so that the damage of device implantation to the brain tissue can be reduced; a flexible material is adopted to be matched with the Young modulus of the brain tissue, so that the nerve scar caused in the in-vivo device implantation process can be reduced, and long-term in-vivo stable recording is realized; and accurate stimulation, in-situ recording and cross-brain-region synchronous recording can be realized through photoelectricsignal interconnection.

The invention relates to the field of online auxiliary evaluation, in particular to a cognitive and brain image data integration evaluation method for Alzheimer's disease. According to the method, firstly, multi-field screening and evaluation of the cognitive function are carried out, the purposes of rapid screening and detailed evaluation of the cognitive function are achieved, then multi-mode magnetic resonance imaging data analysis is carried out, and the multi-mode magnetic resonance imaging data analysis comprises the steps of extracting hippocampus volume and shape analysis, calculating a function connection value of a cognitive-related brain function network, and calculating the cognitive-related brain function network. As an index in the aspect of neuroimaging, the brain function is evaluated. Carrying out cognitive and neural Bayesian joint modeling; and finally, classifying and identifying the specific change mode of the Alzheimer's disease by using supervised learning.

Cerebral cortex thickness estimation in brain magnetic resonance imaging (MRI) is an important technical means for researching brain development and neurodegenerative diseases in neuroimaging. The invention provides a cerebral cortex thickness estimation algorithm based on a three-dimensional Laplace operator, which can accurately capture geometrical morphological characteristics in brain nuclearmagnetic resonance imaging. The method comprises the following steps: firstly, starting from the elimination of a cross overlapping region generated on the surface of a gray matter layer and the surface of a white matter layer, constructing a tetrahedral mesh which reflects the inherent geometrical characteristics of the brain and is matched with MRI; secondly, constructing a three-dimensional Laplace operator by utilizing a geometric constraint relationship of a tetrahedral mesh, and calculating the distribution of a cerebral cortex internal temperature field under a Diels boundary by utilizing a finite element method; then, determining a local isothermal surface, obtaining the gradient line direction of the isothermal surface in the temperature field through a calculation geometry method, and a tetrahedral mesh unit where internal points on a gradient line are located is rapidly locked through a half-half surface data storage structure; and finally obtaining the thickness characteristic information of the cerebral cortex according to the direction and the step length of each gradient line by combining the set gradient step length. According to the method, the morphological structure detection capability of the cerebral cortex can be effectively improved by constructing a high-quality cerebral cortex tetrahedral mesh and determining a high-precision temperature field gradientline.

The invention discloses a method for mining biomarkers based on multi-map neuroimaging data, relates to a method for identifying graphs, and can consider a high-order complementary relationship amongmultiple maps and sample weight information at the same time. Feature analysis is carried out on the neural image data by adopting a multi-map feature selection method based on weight-induced low-ranklearning. According to the method, a first-order neighborhood aggregation mode is adopted, the sum of all connection strengths of each brain region serves as the characteristics of the brain region,the selected characteristics are more stable in a loop iteration mode, finally, the selected characteristics are fused and classified through a multinuclear support vector machine, and thus the Alzheimer's disease diagnosis precision is improved. The method overcomes the defects that in an existing Alzheimer's disease classification technology, sample weight information and multi-map information cannot be considered, and Alzheimer's disease diagnosis and classification errors are prone to occurring.

The present disclosure includes provides methods for assessing resting-state fMRI functional connectivity, resting-state MEGI functional connectivity, and / or task-based spatiotemporal auditory cortical activity latency in a subject to detect, monitor, and / or diagnose Tinnitus, with or without hearing impairment. The present disclosure also provides systems, devices, and methods for diagnosing Tinnitus and / or hearing impairment in a subject. Also provided are systems configured for performing the disclosed methods and computer readable medium storing instructions for performing steps of the disclosed methods.

The invention discloses a neuropathology hub node identification method combining pathology and topological information. The method comprises the following steps: firstly, determining a basic mathematical model of a brain network; then defining a neuropathologic potential and a neuropathologic potential difference, and then defining a neuropathologic potential difference of the whole brain network; constructing an energy function of the neuropathology hub recognition, and determining an optimization method of the neuropathology hub recognition; finally, real nerve image data are preprocessed, and an optimization algorithm is executed to obtain a neuropathology hub; according to the method, the effect of the neuropathology hub in the brain network topological structure and the characteristic expression of the neuropathology load distributed at the neuropathology hub are analyzed in a combined mode, the hub node with the high neuropathology potential is recognized, and the limitation that only topological characteristics of the hub node in the network structure are considered in a traditional hub recognition method is solved.

The invention relates to a construction method and application of a morphological fusion classification index of a neural image marker, and the construction method comprises the following steps: obtaining the structural MRI data of M centers, and extracting the feature data of a brain structure image; independently training the data of each center, respectively establishing a classification model of each center, and obtaining classification models of M centers; for any sample, calculating a classification weight value of the sample in each feature of each model in classification models of all centers, namely an SHAP matrix; and then taking the sample size used during training of each model as a weight, and calculating according to a formula (1) to obtain a single morphological fusion classification index MICI value: Si represents the sample size of the model i, B represents the total number of the features, ai represents the SHAP value of the feature a in the model i, and i = 1-M. The MICI value can well realize identification between mental disease patients and normal people, and has good interpretability, evolutionary property and expandability.

The invention relates to the technical field of artificial intelligence, and provides an image classification method and device, electronic equipment and a storage medium, and the method comprises the steps: determining a to-be-classified neural image; inputting the neural image into a classification model to obtain a classification result of the neural image output by the classification model; the classification model is obtained by training based on a first sample nerve image and a corresponding sample classification result on the basis of a multi-task learning pre-training model, and the multi-task learning pre-training model is obtained by training based on a second sample nerve image and a sample label under each task corresponding to the second sample nerve image on the basis of an unsupervised pre-training model; the unsupervised pre-training model is obtained based on unsupervised training of the third sample neural image. According to the method and device, the electronic equipment and the storage medium provided by the invention, the data labeling cost is saved, the problem of overfitting of the model is avoided, the performance and generalization of the model on an image classification task are improved, and the accuracy of a classification result is improved.

The invention relates to the technical field of artificial intelligence, and provides an image reconstruction method and device, electronic equipment and a storage medium, and the method comprises the steps: determining a to-be-reconstructed neural image, inputting the neural image into an image reconstruction model, and obtaining a reconstruction result of the neural image. An adopted image reconstruction model is obtained by training a target pre-training model through a first sample nerve image and a corresponding sample reconstruction result; and the target pre-training model is obtained by training in three pre-training modes of contrast learning unsupervised pre-training, cross-modal image conversion supervised pre-training and image reconstruction unsupervised pre-training, so that the problem of over-fitting of the model is avoided, the performance and generalization of the model in the aspect of an image reconstruction task are greatly improved, and on the basis, the target pre-training model is obtained by training in the three pre-training modes of contrast learning unsupervised pre-training, cross-modal image conversion supervised pre-training and image reconstruction unsupervised pre-training. The input neural image is reconstructed by using the image reconstruction model, so that the accuracy of the reconstruction result can be greatly improved.

The invention provides a Web-based disturbance of consciousness electroencephalogram data analysis system. A platform mainly comprises two parts, namely an electroencephalogram acquisition experiment platform and a data analysis platform. Firstly, a resting state experiment normal form is selected, electroencephalogram data is collected by using an electroencephalogram cap, and then the data is imported into a rear-end database of the platform for data analysis. The front end performs algorithm selection of electroencephalogram data analysis and presentation of a result interface, and the rear end receives a request of the front end, calls electroencephalogram data in an algorithm database of data analysis, processes the data and returns a result to the front end. According to the disturbance of consciousness (DOC) electroencephalogram data analysis system provided by the invention, a consciousness level evaluation and scientific research platform integrating requirements of medical staff and requirements of scientific researchers can be realized, and rapid conversion of a computer-aided analysis technology based on neural image data on DOC clinical evaluation can be promoted.

The invention belongs to the technical field of nervous system image interaction information processing. The invention discloses a nervous system image interaction information processing system and anervous system image interaction information processing method. The nervous system image interaction information processing system comprises a nervous image acquisition module, a main control module,an image enhancement module, an image feature extraction module, an image compression module, an image classification module, an image encryption module, an image storage module and a display module.The compression efficiency is greatly improved through the image compression module. Compared with an existing method for only performing video compression on a cross section sequence, the direction with the highest spatial correlation or time correlation in the up-down direction, the front-back direction and the left-right direction of the neural image data in the volume or between the volumes can be found, and the compression efficiency is further improved; and the high-performance standard video encoder is adopted to perform nerve image compression, so that the compressed nerve image data can be decoded conveniently, and the popularization is easy.

The invention relates to the technical field of neural image processing, and particularly discloses a construction method and system for integrating a neural image data analysis environment, and the method comprises the steps: constructing a docker mirror image according to a preset Dockerfile; creating a container according to the docker mirror image; in the created container, cancelling the access control of the host machine by controlling the cancelling instruction, so that the graphical interface in the container can be correctly displayed on the host machine; performing backup storage on the docker mirror image according to a preset storage format and a preset storage path; and decompressing the constructed and backed-up docker mirror image according to a decompression instruction, and transmitting the decompressed docker mirror image into a loading instruction through a pipeline as standard output for mirror image loading and the like. According to the invention, the construction and deployment process of the neural image data analysis and processing environment is simplified, and the purposes of one-button construction, operation, backup and deployment of the neural image data processing environment which is stable and repeatable and can be used for machine learning and deep learning are achieved.

The invention provides a psychological physical test method and equipment for negative emotion of a drug addiction patient and a medium, and belongs to the technical field of addiction evaluation. The problem that existing test standards are not uniform is solved. The psychological physical test method for the negative emotion of the drug addiction patient comprises the following steps: S1, inputting external stimulation; S2, testing negative emotion; S3, outputting behavior reaction; wherein the step S2 comprises the following steps: S2.1, testing negative emotion behavior; S2.2, carrying out a negative emotion scale test; and S2.3, carrying out a nerve image test. The method has the advantage of unified test standard.

Systems and methods for decoding chord information from brain activity are disclosed, comprising: acquiring raw data of their brain activity when one or more subjects listen to music tagged with chord tags; extracting a brain activity mode from the original brain activity data; coupling the brain activity pattern and the music data in time to form training data of the decoding model; training a decoding model; optionally, the trained decoding model is trimmed by marking and using a small amount of originally unmarked brain activity data; acquiring a second batch of original brain activities from the subject via functional neural imaging in the various psychological music activities; and mapping the second batch of brain activities to corresponding chord information.

The invention relates to an intracranial operation navigation system based on magnetic resonance imaging. The system includes a preoperative planning module, a positioning conversion module, a three-dimensional neuroimaging visualization module, a three-dimensional neuroimaging rapid segmentation module, and an intraoperative guidance module. The operation navigation system of the invention can apply a computer aided technology and a magnetic resonance imaging technology to an intracranial surgery process, effectively helps doctors to perform preoperative surgical planning and intraoperative surgical positioning and guidance, and assists surgical operation, thereby reducing difficulty of surgery and treatment accuracy and effects are improved.

The invention belongs to the technical field of neuroimaging, and particularly relates to a research method for a teenager muscular spasm epilepsy multi-brain-region co-activation mode. The method comprises the following steps: firstly, analyzing the grey matter volume of a JME patient by utilizing voxel-based morphological analysis (VBM); it is assumed that a brain region with gray matter volume atrophy is a key brain region causing disease generation, so that the brain region with gray matter volume atrophy is used as a region of interest, and brain dynamic activity changes of a JME patient and a healthy control group in the selected region of interest are analyzed through a co-activation mode. Therefore, the difference between the dynamic brain activities of the JME patient and the normal person in the resting state can be searched.

The invention provides a modular neurocognitive function test method for drug addiction evaluation, and belongs to the field of addiction assessment. The problem that existing evaluation standards are not uniform is solved. The modular neurocognitive function test method comprises the following steps of S1, inputting external stimulation; S2, carrying out modularized neurocognitive function test; S3, carrying out behavior reaction output, wherein the step S2 comprises the following steps of S2.1, executing a function test; S2.2, carrying out negative emotion testing; S2.3, carrying out a reward motivation test; the step S2.1 comprises the following sub-steps of S2.1. 1, paying attention to a control test; S2.1.2, carrying out a reaction inhibition test; S2.1. 3, carrying out a plan processing test; S2.1. 4, carrying out a work memory test; S2.1. 5, carrying out a pre-estimation decision test; the step S2.2 comprises the following steps of S2.2. 1, carrying out a negative emotion behavior test; S2.2. 2, carrying out a negative emotion scale test; S2.2. 3, carrying out a negative emotion nerve image test; and the step S2.3 comprises the following steps of S2.3. 1, performing a reward motivation behavior test; S2.3. 2, carrying out a reward motivator scale test; and S2.3. 3, carrying out a reward motivator nerve image test. The method has the advantage of unified standard.

The invention discloses a biomarker mining method based on multi-atlas neuroimaging data, relates to a method for identifying graphics, which can simultaneously consider high-order complementary relationships between multiple atlases and sample weight information, and adopts weight-based low-rank learning Multi-atlas feature selection method for feature analysis of neuroimaging data. This method adopts the method of first-order neighborhood aggregation, takes the sum of all connection strengths of each brain region as the feature of the brain region, and uses the loop iteration method to make the selected features more stable, and finally uses the multi-core support vector machine to The selected features are fused and classified to improve the diagnostic accuracy of Alzheimer's disease. The present invention overcomes the defect that the existing Alzheimer's disease classification technology cannot consider sample weight information and multi-atlas information, and is easy to misclassify Alzheimer's disease diagnosis.

The present invention is a construction method and application of neuroimaging marker morphological fusion classification index. The construction method includes the following contents: obtaining structural MRI data of M centers, extracting brain structural image characteristic data; performing independent training with the data of each center, respectively Establish the classification model of each center and obtain the classification model of M centers; for any sample, in the classification model of all centers, calculate the classification weight value of the sample in each feature of each model, that is, the SHAP matrix; then use each The sample size used in the training of each model is the weight, and a single morphological fusion classification index MICI value is obtained by calculating according to formula (1): where Si represents the sample size of model i, B represents the total number of features, and a i Represents the SHAP value of feature a in model i, i=1~M. The MICI value can well realize the discrimination between patients with mental illness and normal people, and has good interpretability, evolution and scalability.

The processing method of multi-modal brain neuroimaging features of the present invention relates to the image preprocessing for the extraction of image features or characteristics for identifying graphics. Firstly, the multi-modal neuroimaging feature selection method with sample weight and low-rank constraints is used for multi-modal neuroimaging. Perform feature selection on the data of the modality to obtain the low-dimensional feature matrix, calculate the kernel matrix of each modality, obtain the low-dimensional feature matrix, calculate the kernel matrix of each modality, and then fuse the kernel matrices of different modalities into one kernel matrix, thereby selecting more discriminative biomarker features, and using multi-core support vector machines to predict and classify new sample cases of Alzheimer's disease, which overcomes the existing biomarker used in the prior art The characteristics will cause harm to the subject, and the patient's diseased brain region cannot be found by using only one brain image feature data or insufficient multimodal brain neuroimaging feature data, and the brain images used The features have no medically interpretable flaws.

The present invention relates to a system and method for training an artificial intelligence (AI) based neural imaging system for evaluating and replicating actuarial calculation patterns of known valuation systems. In particular, present inventions disclose evaluating, using neural imager, output from a data generation unit and output from the actuarial assumption generation unit with the output from a valuation system to generate at least an image model replicating the actuarial calculations of the valuation system using neural networks.