GB2578929A - Automatic brain disorder tendency screening using statistical features of MEG data - Google Patents
Automatic brain disorder tendency screening using statistical features of MEG data Download PDFInfo
- Publication number
- GB2578929A GB2578929A GB1900786.3A GB201900786A GB2578929A GB 2578929 A GB2578929 A GB 2578929A GB 201900786 A GB201900786 A GB 201900786A GB 2578929 A GB2578929 A GB 2578929A
- Authority
- GB
- United Kingdom
- Prior art keywords
- brain
- data set
- meg data
- meg
- statistical features
- 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.)
- Granted
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
- A61B5/4088—Diagnosing of monitoring cognitive diseases, e.g. Alzheimer, prion diseases or dementia
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4058—Detecting, measuring or recording for evaluating the nervous system for evaluating the central nervous system
- A61B5/4064—Evaluating the brain
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
- A61B5/4094—Diagnosing or monitoring seizure diseases, e.g. epilepsy
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Neurology (AREA)
- Physiology (AREA)
- Mathematical Physics (AREA)
- Psychology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Developmental Disabilities (AREA)
- Neurosurgery (AREA)
- Child & Adolescent Psychology (AREA)
- Hospice & Palliative Care (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
A brain disorder, e.g. epilepsy or Alzheimer’s, is detected by identifying MEG data sets 214 with the brain disorder 220. This is done by training 208 a Support Vector Machine (SVM) by segmenting MEG data from different brain regions (such as left temporal etc.) into possibly one-minute epochs. For each multi-channel segment of a brain region, the signals are concatenated and a set of statistical features is extracted from them, for both healthy brain data and data representing the brain disorder. The extracted statistical features may include maximum or minimum or comprise other parameters such as standard deviation or kurtosis. There may for example be eight statistical features from each of eight brain regions, totalling 64. Following the training phase 202, the trained model 210 is used to classify incoming feature vectors in the diagnosis phase 212, to distinguish e.g. healthy from epileptic data.
Description
AUTOMATIC BRAIN DISORDER TENDENCY SCREENING USING
STATISTICAL FEATURES OF MEG DATA
FIELD OF THE INVENTION
[0001] The present disclosure relates generally to assess a subject at risk to have or acquire a brain disorder. More particularly, relates to systems and methods of detecting subjects with a brain disorder using machine training technology.
BACKGROUND OF THE INVENTION
[0002] A neurological disorder is any disorder of the nervous system causing structural, biochemical or electrical abnormalities in the brain, spinal cord or other nerves. A neurological disorder may cause a range of symptoms, which include paralysis, muscle weakness, poor coordination, loss of sensation, seizures, confusion, pain and altered levels of consciousness.
[0003] Epilepsy is a neurological brain disorder that affects millions of people worldwide, strikes at any age, produces a variety of behaviors, and is considered as the second most serious disease after stroke (Schulze-Bonhag, 2010; WHO report, 2012). Different technologies have been used to monitor brain activities, such as electroencephalogram (EEG), computerized tomography (CT), and magnetic resonance imaging (MR1). Due to the heterogeneity in conductivity within head tissues, EEG signals suffer from degradation. On the contrary, the magnetic signals associated with the electric currents, detected by magnetoencephalography (MEG) (Cohen, 1968; Ham, 1993), suffer limited degradation. [0004] Epilepsy largely depends on clinical diagnosis and on clinical history provided by the patient and witnesses. All the additional investigational tools so far do not confirm or dispute the diagnosis, or prove the concept of tendency. EEG is the main ancillary test epileptolog sts use is value is great when the patient has a clinical history of seizures, but if there is no supporting history, then its value is limited to judge the person's tendency to develop seizures. Tendency to have or acquire certain diseases is crucial in the prevention and management of different diseases. For neurological brain disorders and specifically epilepsy, still to date, the field lacks a reliable method of ascertain the diagnosis, assessment or tendency in population at risk.
[00051Aceordingly, there remains the need for a reliable and fast, yet sensitive and specific tool for detecting and assessing subjects at risk to have or acquire brain disorders. SUMMARY OF THE INVENTION [00061 The Summary of this invention is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter. The claimed invention relates to a system and method supporting a computerized information system. More specifically, the claimed invention relates to a system and method for computer-based information technology to detect and diagnose a subject with a brain disorder using statistical features of MEG data and SVM in a computer-based system.
[00071The claimed invention provides a solution in order to overcome a problem specifically arising in the realm of diagnosing brain disorders and more specifically neurological brain disorders using epilepsies as an example. The claims provide a high specific yet sensitive, system and method for diagnosing a subject with a brain disorder. The claimed invention overcomes the limitations of current diagnosing and detection tools of brain disorders, and more specifically neurological brain disorders using epilepsies as an example, and provides other benefits that will become clear to those skilled in the art from the foregoing description.
[00081 Accordingly, in one aspect, the embodiments of the present invention provide for a non-transitory computer readable medium storing specific computer-executable instructions for detecting a brain disorder that, when executed by a processor, cause a computer system to automatically at least: (a) develop a first training set for a support vector machine (SVM) by: collecting a first Magnetoencephalography (MEG) data set of different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the brain disorder; (b) develop a second training set for the SVM by: collecting a second MEG data set of the different brain regions that represent a healthy brain; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (c) train the support vector machine to classify the state of an MEG data set based on differentiating between the first and second training sets; and (d) screen a set of MEG data by the SVM by: collecting an MEG data set of the different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets with the brain disorder. In these embodiments, steps (a)-(c) constitute a training phase for the SVM. In some embodiments, the non-overlapping epochs are 1 minute epochs.
[0009] In these embodiments, the brain disorder s op lepsies, multiple sclerosis, Alzheimer's disease, autism, Parkinson, ADHA, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain. In some embodiments, the brain disorder is epilepsies.
[00010] In some embodiments, the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
[00011] In the embodiments of the present invention, the statistical features are selected from the group comprising, but not limited to: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
[00012] In another aspect, the present invention further provides for a system for detecting a brain disorder, the system comprising: (a) a computer server that stores MEG data sets for a plurality of subjects including subjects with the brain disorder and healthy subjects; (b) One or more computer storage media having computer-usable instructions that, when used by one or more computing devices, cause the one or more computing devices to perform a method for brain disorder risk assessment, the method comprising: (i) developing a first training set for a support vector machine (SVM) by: collecting a first MEG data set of different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the subjects of the brain disorde * (ii) developing a second training set for the SVM bv: collecting a second MEG data set of the different brain regions that represent healthy subjects; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (iii) training the support vector machine to classify the state of an MEG data set based on differentiating between the first and second training sets; and (iv) screening a set of MEG data by the SVM by: collecting an MEG data set of different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets relate to subjects with the brain disorder. In these aspects of the present invention, steps (i)-(iii) constitute a training phase for the SVM.
[00013] In aspects of the present invention, the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain. In some aspects, the brain disorder is epilepsies.
[00014] In some aspects, the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head. In some aspects of the present invention, the statistical features arc selected from the group comprising, but not limited to: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
[00015] Additionally, the embodiments of the present invention provide for a computer-implemented method for in vitro detecting a subject with a brain disorder, the method comprising: (a) developing a first training set for a support vector machine (SVM) by: collecting a first MEG data set of the different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the brain disorder; (b) developing a second training set for the SVM by: collecting a second MEG data set of different brain regions from that represent a healthy brain; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (c) training the SVM to classify the state of an MEG data set based on differentiating between the first and second training sets; and (d) screening a set of MEG data by the SVM by: collecting an MEG data set of the different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets with the brain disorder. In these embodiments, steps (a)-(c) constitute a training phase for the SVM.
[00016] In the embodiments of the present invention, the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain. in some embodiments, the brain disorder is epilepsies.
[00017] In some embodiments, the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head. In some embodiments, the statistical features are selected from the group comprising, but not limited to: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
[000181 Hereinafter the different embodiments and aspects of the present invention is described in detail, however the scope of the present invention should not be restricted to these descriptions, even with the addition to the following examples as appropriate without departing from the spirit of the present invention it may change implementation.
BRIEF DESCRIPTION OF THE DRAWINGS
[00019] Figure 1 depicts a diagram illustrating an operating environment suitable for practicing an embodiment of the invention.
[00020] Figure 2 provide a flow diagram illustrating the overall model scheme in accordance with the embodiments of the present invention_ [00021] Figure 3 depicts a block diagram of a high level design or architecture for an embodiment of the invention showing the training phase [00022] Figure 4 depicts a block diagram of a high level design or architecture for an embodiment of the invention showing the diagnosing phase.
[00023] Figure 5 shows a diagram illustrating the performance of the diagnostic algorithm.
DETAILED DESCRIPTION OF THE INVENTION
[00024] It is an object of the present invention to provide superior tools for diagnosing and detecting brain disorders, using epilepsy as an example, with high specificity and sensitivity. The embodiments of the present invention provide a system, method, or set of instructions embodied on one or more computer-readable media to diagnose, detect, assess the tendency to have or acquire brain disorders for a subject at risk. More particularly, the embodiments of the present invention provide for a system, method, or set of instructions embodied on one or more computer-readable media to classify MEG signals based on statistical features for classifying MEG signals into brain disorders and healthy subjects. In some embodiments, the brain disorder is a neurological brain disorder. In other embodiments, the brain disorder is epilepsy.
[00025] Epilepsy is a neurological brain disorder that affects millions of people worldwide, strikes at any age, produces a variety of behaviors, and is considered as the second most serious disease after stroke (Schulze-Bonhag, 2010; WHO report, 2012). Different technologies have been used to monitor brain activities, such as electroencephalogram (EEG), computerized tomography (CT), and magnetic resonance imaging (MRI). The primary tool for epilepsy diagnosis is EEG, which measures electrical brain activity. For epilepsy diagnosis, epileptologists spend great effort reading the long EEG recordings.
[00026] Researchers have invested a great deal of effort in epilepsy-related studies, especially seizure detection and prediction based on processing and analyzing EEG signals (Acharya, 2013; Gadhoumi, 2015). Nevertheless, due to the heterogeneity in conductivity within head tissues, EEG signals suffer from degradation. On the contrary, the magnetic signals associated with the electric currents, detected by MEG (Cohen, 1968; Ham 1993), suffer limited degradation. The weak magnetic signals outside the head are measured using a superconducting quantum interference device (SQUID) (Cohen, 1972). No electrodes need to be attached to the head, so it is possible to screen large numbers of patients quickly and easily. MEG's helmet contains more than 300 sensors, which cover the whole head with high-resolution capabilities. Clinically, MEG is used for pre-surgical evaluation and vascular malfunction (De Tiege, 2012; Kharkar, 2015; and Ossenblok, 2007).
[00027] Diagnosis of epilepsy depends in large on clinical diagnosis and based mainly on clinical history provided by the patient and witnesses. So far, no additional investigational tools are available to confirm or dispute the diagnosis, nonetheless to prove the concept of tendency. EEG is the main ancillary test which epileptologists use, its value is great when the patient has a clinical history of seizures, however if there are no supporting history, its value diminishes to evaluate the tendency in a person at risk of developing seizures. For disease prevention and management, it's crucial to access the tendency of a subject to have or acquire that disease. This can be applied to epilepsy, but we arc still lacking a reliable method of ascertain the diagnosis or tendency in population at risk.
[00028] The embodiments of the present invention provide for a new model that processes data obtained from a diagnostic tool which records signals obtained from patients (data) Using this data, the algorithm diagnose brain disorder, using epilepsy as an example, with a very high confidence rate in a short time period. Currently, the data obtained from the diagnostic tool is analyzed manually for the diagnosis of epilepsy; a laborious activity with high error rate. The embodiments of the present invention provide method and system which automatically analyze, diagnose, assess, or evaluate the tendency of subjects at risk to have or acquires a brain disorder using statistical features of MEG data and SVM.
[00029] The aspects of the present invention provide for a system, method, or set of instructions embodied on one or more computer-readable media to classify MEG signals based on statistical features for classifying MEG signals into brain disorders and healthy subjects. For the purpose of clarity, Computer-readable media include media implemented for storing information. Examples of stored information include computer-useable instructions, data structures, program modules, and other data representations. By way of example, and not limitation, computer-readable media may comprise computer storage media. Computer-readable media include both volatile and nonvolatile media, removable and non-removable media, and contemplate media readable by a database, a switch, and various other network devices. Media examples include hardware memory devices such as RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile discs (DVD), holographic media or other optical disc storage, magnetic cassettes, magnetic tape, magnetic disk storage, and other data storage devices. These technologies can store data momentarily, temporarily, or permanently. As used herein, computer-readable media do not include signals per se.
[00030] In a first aspect, a computer-implemented method is provided for in vitro detecting a subject with a brain disorder. The method comprises two phases: training phase and diagnosis phase. in the training phase, three tasks are performed; data segmlentation, feature extraction, and training the classifier. in diagnosis phase, the MEG data is segmented and the features are extracted as in phase one. Then the trained classifier is tasked to diagnose whether the incoming segment is related to healthy or epileptic subject.
[00031] in a second aspect, a system is provided for detecting a subject with a brain disorder, wherein the system comprises: (1) a computer server that stores MEG data sets for a plurality of subjects including subjects with the brain disorder and healthy subjects; (2) one or more computer storage media having computer-usable instructions that, when used by one or more computing devices, cause the one or more computing devices to perform a method for brain disorder risk assessment using machine training for a SVM.
[00032] in a third aspect, a non-transitory computer readable medium storing specific computer-executable instructions for detecting a subject with a brain disorder that, when executed by a processor, cause a computer system to automatically execute a method of brain disorder risk assessment performing a machine training for a SVM to classify MEG signals to a brain disorder or healthy subjects using statistical features.
[00033] Having briefly illustrated embodiments of the present invention, an exemplary operating environment suitable for use in implementing embodiments of the present invention is described below. Referring to the drawings in general, and initially to Figure I in particular, an exemplary computing system environment, for instance, a medical information computing system, on which embodiments of the present invention may be implemented is illustrated and designated generally as reference numeral 100. It will be understood and appreciated by those of ordinary skill in the art that the illustrated medical information computing system environment 100 is merely an example of one suitable computing environment and is not intended to suggest any limitation as to the scope of use or functionality of the invention. Neither should the medical information computing system environment 100 be interpreted as having any dependency or requirement relating to any single component or combination of components illustrated therein.
[00034] The present invention is a special computing environment that can leverage well-known computing system environments or configurations. Examples of well-known computing systems, environments, and/or configurations that may be suitable for use with the present invention include, by way of example only, personal computers, server computers, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above-mentioned systems or devices, and the like.
[00035] The present invention may be described in the context of computer-executable instructions, such as program modules, being executed by a computer.
Exemplary program modules include, but arc not limited to, routines, programs, objects, components, and data structures that perform particular tasks or implement particular abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in local and/or remote computer storage media including, by way of example only, memory storage devices.
[00036] With continued reference to Figure 1, the exemplary computing system environment 100 includes a general-purpose computing device in the form of a server 102. Components of the server 102 may include, without limitation, a processing unit, internal system memory, and a suitable system bus for coupling various system components, including database cluster 104, with the server 102. The system bus may be any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, and a local bus, using any of a variety of bus architectures. By way of example, and not limitation, such architectures include industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronic Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus, also known as Mezzanine bus.
[00037] The server 102 typically includes, or has access to, a variety of computer-readable media, for nstmce, database cluster 104. Computer-readable media can be any available media that may be accessed by server 102, and includes volatile and nonvolatile media, as well as removable and non-removable media. By way of example, and not limitation, computer-readable media may include computer storage media and communication media. Computer storage media may include, without limitation, volatile and nonvolatile media, as well as removable and non-removable media implemented in any method or teelmology for storage of information, such as computer-readable instructions, data structures, program modules, or other data. In this regard, computer storage media may include, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVDs) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage, or other magnetic storage device, or any other medium which can be used to store the desired information and which may be accessed by the server 102. Communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transport mechanism, and may include any information delivery media. As used herein, the tent "modulated data signal" refers to a signal that has one or more of its attributes set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above also may be included within the scope of computer-readable media.
[00038] The computer storage media discussed above and illustrated in Figure 1, including database cluster 104, provide storage of computer-readable instructions, data structures, program modules, and other data for the server 102.
[00039] The server 102 may operate in a computer network 106 using logical connections to one or more remote computers 108. Remote computers 108 may be located at a variety of locations in a registration environment, for example, but not limited to, education, government, financial, clinical laboratories, entities and other in person settings, billing and financial offices, administration settings, etc. The remote computers 108 may also be physically located in nontraditional environments so that the entire community may be capable of integration on the network. The remote computers 108 may be personal computers, servers, routers, network PCs, peer devices, other common network nodes, or the like, and may include some or all of the components described above in relation to the server 102. The devices can be personal digital assistants or other like devices.
[00040] Computer networks 106 comprise local area networks (LANs) and/or wide area networks (WANs) Such networking environments are commonplace in offices, enterprise-wide computer networks, intranets, and the Internet. When utilized in a WAN networking environment, the server 102 may include a modem or other means for establishing communications over the WAN, such as the Internet. In a networked environment, program modules or portions thereof may be stored in the server 102, in the database cluster 104, or on any of the remote computers 108. For example, and not by way of limitation, various application programs may reside on the memory associated with any one or more of the remote computers 108. For example, an application program 110 may reside on, and be executed by, server 102 or another server, in which case remote computer 108 would access application 110 remotely. it will be appreciated by those of ordinary skill in the art that the network connections shown are exemplary and other means of establishing a communications link between the computers (e.g., server 102 and remote computers 108) may be utilized.
[00041] In operation, a user may enter commands and information into the server 102 or convey the commands and information to the server 102 via one or more of the remote computers 108 through input devices, such as a keyboard, a pointing device (commonly referred to as a mouse), a trackball, or a touch pad. Other input devices may include, without limitation, microphones, satellite dishes, scanners, or the like. Commands and information may also be sent directly from a remote healthcare device to the server 102. In addition to a monitor, the server 102 and/or remote computers 108 may include other peripheral output devices, such as speakers and a printer.
[00042] Although many other ntemal components of the server 102 and the remote computers 108 are not shown, those of ordinary skill in the art will appreciate that such components and their interconnection are well known. Accordingly, additional details concerning the internal construction of the server 102 and the remote computers 108 are not further disclosed herein.
[00043] Turning now to Figure 2, block diagram 200 depicts a high-level system design or architecture for the embodiments of the present invention. The system 200 comprises two major phases "Training Phase" 202 and "Diagnosis Phase" 212. The "Training Phase" phase 202 is designed to perform machine learning or training for SVM by providing "Training MEG signals" 204 for healthy and brain disorder subjects, then perform "Statistical-based feature extraction" 206, and finally "Classifier Training" 208.
1000421 The "Training Phase" 202 is a machine learning or training phase that is done entirely via a script. Like most machine learning techniques, the program must be fed with datasets to analyze and to learn from. The "Training Phase" 202 is required to build a "Trained Model" 210, which is tasked to undergo the "Diagnosis Phase" 212. in the "Diagnosis Phase" 212, the "Trained Model" 210 performs the "Diagnose MEG signals" 214 step, then "Statistical-based feature extraction" 216, and finally perform "Diagnosis" 218 to classify "Healthy/Epileptic" 220 subjects.
[00044] Turning now to Figure 3, block diagram 300 depicts a more in detail design or architecture of the different steps of the "Training Phase" 302. A first training set 304 with MEG signals data of subjects with a brain disorder is provided to SVM. In the first training set 304, the first MEG signals data 306 is collected. Then, the multi-channel MEG signal data is segmented 308 into non-overlapping epochs. For each multi-channel segment of a brain region, the channel-segments are concatenated 310, then, statistical features are extracted 312 from the concatenated signals. Subsequently, SVM is trained with statistical features vectors extracted from brain disorder MEG data 314.
[00045] Similarly, SVM is provided with a second training set 316 with MEG signals data of healthy subjects. In the second training set 316, the second MEG signals data of healthy subjects 318 is collected. Then, the multi-channel MEG signal data is segmented 320 into non-overlapping epochs. For each multi-channel segment of a brain region, the channel-segments are concatenated 322, then, statistical features are extracted 324 from the concatenated signals. Then, SVM is trained with statistical features vectors extracted from healthy MEG data 326. in collection, classifier training to automatically distinguish between brain disorder and healthy subjects is performed to build the "Trained Model-210 [00046] Turning now to Figure 4, block diagram 400 depicts a more in detail design or architecture of the different steps of the "Diagnosis Phase" 400. The SVM model is trained to collect MEG signal data 402, perform segmentation 404, concatenate the segmented MEG signal data 406, and then extract statistical features 408 from the concatenate MEG signal data. The trained SVM model classifies any incoming feature vector to either healthy or epileptic segment 410, hence the trained model is able to detect and classify subjects as healthy or with a brain disorder 412.
[00047] it is an object of the present invention to provide a system and computer-implemented method for in vitro detecting a brain disorder in the embodiments of the present invention, the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain. In some embodiments, the brain disorder is epilepsies. Aspects of the technology described herein are directed to, among others things, a proposed methodology consists of two phases: training phase and diagnosis phase. In the training phase, three tasks are performed; data segmentation, feature extraction, and training the classifier. In diagnosis phase, the MEG data is segmented and the features are extracted as in phase one. Then the trained classifier is tasked to diagnose whether the incoming segment is related to healthy or epileptic subject. Figure 2 presents the overall scheme of the approach.
1000481 The MEG data are collected from subjects in resting-state supine position at a sampling frequency of 1000 Hz and bandpass of 0.03 to 330 Hz using 306 channels. In some embodiments of the present invention, the MEG sensors were separated into eight non-overlapping sets roughly covering the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
[00049] For feature extraction, first, the multi-channel MEG signal is segmented into non-overlapping 1-min epochs. For each multi-channel segment of a brain region, the channel-segments are concatenated. Then, eight statistical features (maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness. The mean and median are the most used measures for central tendency) are extracted front the concatenated signals. The total number of features extracted from all brain regions is 64 (eight features from each brain region). Equation 1 is used to obtain the mean (t), where Xi is the ith data point of the segment and N is of the segment length. Standard deviation and the interquartile range are used to measure statistical dispersion of a segments data points. Equation 2 is used to calculate the standard deviation (u). The interquartile range (11)R) is the difference between the median of the lower half data points (Q1) and that of the upper half data points (Q3). Kurtosis measures the eurviness of the data distribution, whether it is peaked or flat. Skewness measures the symmetry of the data distribution around the mean. Equations 3 and 4 are used to calculate the kurtosis and skewness, respectively (XTST/SEMATECH, 2016).
[00050] The SVM is a supervised machine learning algorithm that was introduced in the 1990s, and it became one of the most popular machine learning algorithms (Boser, 1992; Vapnik, 2000). It has been applied successfully as a classifier in different areas (Wang, 2005). Chan et al. and Ince et al. applied SVM on MEG recordings for word decoding and schizophrenia classification (Chan, 2011; Ince, 2008). In the training phase, SVM with a Radial Basis Function (RBF) kernel is trained with statistical features vectors extracted from epileptic and healthy MEG data. In the diagnosis phase, the trained SVM model is utilized to classify any incoming feature vector to either healthy or epileptic segment.
[00051] The embodiments of the present invention provide for a computer-implemented method for vitro detection of a subject with a brain disorder, the method comprising: (a) developing a first training set for a support vector machine (SVM) by: collecting a first MEG data set of the different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the brain disorder; (b) developing a second skewneo = training set for the SVM by: collecting a second MEG data set of different brain regions from that represent a healthy brain; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (c) training the SVM to classify the state of an MEG data set based on differentiating between the first and second training sets; and (d) screening a set of MEG data by the SVM by: collecting an MEG data set of the different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets with the brain disorder. in these embodiments, steps (a)-(c) constitute a training phase for the SVM.
[00052] In these embodiments, the brain disorder is a neurological brain disorder.
In these embodiments, the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Slogren's syndrome, chronic alcoholism, or facial pain in some embodiments, the brain disorder is epilepsies.
[00053] In some embodiments, the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
[00054] In some embodiments of the present invention the statistical features are selected from the group comprising, but is not limited to: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
[000551 In other aspects of the present invention, system for detecting subjects with a brain disorder, the system comprising: (a) a computer server that stores MEG data sets for a plurality of subjects including subjects with the brain disorder and healthy subjects; (b) One or more computer storage media having computer-usable instructions that, when used by one or more computing devices, cause the one or more computing devices to perform a method for brain disorder risk assessment, the method comprising: (i) developing a first training set for a support vector machine (SVM) by: collecting a first MEG data set of different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the subjects of the brain disorder; (ii) developing a second training set for the SVM by: collecting a second MEG data set of the different brain regions that represent healthy subjects; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (iii) training the support vector machine to classify the state of an MEG data set based on differentiating between the first and second training sets; and (iv) screening a set of MEG data by the SVM by: collecting an MEG data set of different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistic& features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets relate to subjects with the brain disorder. In these aspects. steps (i)-(iii) constitute a training phase for the SVM.
[00056] In some aspects, the brain disorder is a neurological brain disorder. In these aspects, the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain. In some aspects, the brain disorder is epilepsies.
[00057] In some aspects, the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
[00058] In other aspects, the statistical features arc selected from the group comprising, but is not limited to: maximum value, minimum value, mean, standard deviation, median, intcrquartilc range, kurtosis. and skewness.
[00059] In different embodiments of the present invention, a non-transitory computer readable medium storing specific computer-executable instructions for detecting subjects with a brain disorder, that, when executed by a processor, cause a computer system to automatically at least: (a) develop a first training set for a support vector machine (SVM) by: collecting a first Magnetoencephalography (MEG) data set of different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the brain disorder; (b) develop a second training set for the SVM by: collecting a second MEG data set of the different brain regions that represent a healthy brain; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (c) train the support vector machine to classify the state of an MEG data set based on differentiating between the first and second training sets; and (d) screen a set of MEG data by the SVM by: collecting an MEG data set of the different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets with the brain disorder. In these embodiments, steps (a)-(c) constitute a training phase for the SVM. in some embodiments, the non-overlapping epochs are I minute epochs.
[00060] In some embodiments, the brain disorder is a neurological brain disorder.
In these aspects, the brain disorder is epilepsics, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain. In some aspects, the brain disorder is epilepsies.
[00061] In some embodiments, the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RE), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
[00062] In other embodiments, the statistical features are selected from the group comprising, but not limited to: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
[00063] The embodiments of the present invention provide for a novel model that processes data obtained from a diagnostic tool which records signals obtained from subjects at risk. Using this data, the novel model of the present invention diagnoses brain disorder, using epilepsy as an example, with a very high confidence rate in a short time period. The embodiments of the present invention provide a method and system which automatically analyze, diagnose, assess, or evaluate the tendency of subjects at risk to have or acquire a brain disorder using statistical features of MEG data and SVM.
[00064] The subject matter of the embodiments of the present invention is described with specificity herein to meet statutory requirements. However, the description itself is not intended to limit the scope of this patent. Rather, the inventors have contemplated that the claimed subject matter might be embodied in other ways, to include different elements or combinations of elements similar to the ones described in this document, in conjunction with other present or future technologies. Terris should not be interpreted as implying any particular order among or between various steps herein disclosed unless and except when the order of individual steps is explicitly described.
EXAMPLES
Example 1: Automated Epileptic Tendency Screening Using Statistical Features of MEC Data and SVM.
[00065] The proposed methodology consists of two phases: training phase and diagnosis phase. In the training phase, three tasks are performed; data segmentation, feature extraction, and training the classifier. In diagnosis phase, the MEG data is segmented and the features are extracted as in phase one. Then the trained classifier is tasked to diagnose whether the incoming segment is related to healthy or epileptic subject. Figure I presents the overall scheme of the approach.
Clinical MEG Data [00066] The proposed methodology was tested on high-quality MEG data recorded at National Neuro Institute (NNI), King Fahad Medical City (KFMC), Riyadh, Saudi Arabia, using Elckta Neuromag in a shielded room. The MEG data were collected from subjects in resting-state supine position at a sampling frequency of 1000 Hz and bandpass of 0.03 to 330 Hz using 306 channels. MEG sensors were separated into eight non-overlapping sets roughly covering the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head. Each set contained 26 gradiometers, except LO and RO sets, which contained 24 channels each. A total of 64 subjects' (32 healthy and 32 epileptic patients) MEG data recordings were used.
Results and Discussion [00067] The suggested screening approach was applied on MEG data of 64 subjects with a total number of 596 and 613 minutes for healthy and epileptic subjects, respectively. A four-fold cross-validation strategy is used. Each fold contains segments from eight healthy subjects and eight Non-healthy subjects. After the feature extraction phase, SVM is used to classify the segments of each fold after training the model on the segments of the other three folds. For evaluating the performance of the proposed method, we used two metrics: sensitivity and specificity. Sensitivity represents the ratio of number of times the classifier makes correct positive decisions (i.e., detects epileptic segments) to the total number of positive decisions [00068] While specificity, is the ratio of number of times the classifier makes correct negative decisions (i.e., detects healthy segments) to the total number of negative decisions (Khalid, 2015). The proposed method achieved an average sensitivity and specificity of 99.35% and 95.47%, respectively. thrall et al. (Khalid, 2015) method, which uses the whole subject's data for feature extraction and LDA as a classifier (quadratic), achieved an average sensitivity and specificity of 100% and 90.63%, respectively, on the same dataset (64 subjects).
[00069] Figure 5 shows the rate of correctly classified segments for all 64 subjects.
The bars towards the top illustrate the epileptic subjects and the bars towards the bottom illustrate the healthy subjects. We can see that the proposed method identifies the epileptic subjects' segments with almost 100% accuracy_ However, some segments of the healthy subjects are classified as epileptic. Only two healthy subjects, 7 and 24, classified less than 85?/0 of their segments correctly (67% and 60%), while all subject 6's segments were classified as an epileptic subject. Epileptologists, in KFMC, reviewed their recordings and reported the following. Subject 24 had MEG electrodes contaminated with artifacts, and subject 7 had an epileptic waveform, which in this case can be considered incidentals findings.
[00070] In case of subject 6, they found he has the well-described normal variant (lambda), which is a normal waveform found in awake subjects with visual intention toward any object (Brigo, 2011).
[00071] The present invention provides an automated brain disorder, using epilepsies as an example, diagnostic tool based on statistical features extracted from MEG data is proposed. MEG is gaining popularity worldwide in analyzing brain activities. The MEG is segmented into 1-minute non-overlapping segments, and then eight statistical features are extracted. The SVM is trained on features vectors and then tasked to classify the incoming features vectors. A four-fold cross-validation strategy is adopted in this work. The proposed algorithm was tested on MEG real data of 64 subjects (32 epileptic and 32 healthy subjects) and achieved a sensitivity of 99.35%, a specificity of 95.47%. This approach is a valuable tool for epileptologists, which can help them in the epilepsy, and brain disorders generally, diagnosis process.
EQUIVALENTS
[00072] While the invention has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth and as follows in the scope of the appended claims.
[00073] Those skilled in the art will recognize, or be able to ascertain, using no more than routine experimentation, numerous equivalents to the specific embodiments described specifically herein. Such equivalents are intended to be encompassed in the scope of the following claims.
INCORPORATION BY REFERENCE
[00074] All patents and publications referenced herein are hereby incorporated by reference in their entireties. The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be constmed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention.
[00075] As used herein, all headings are simply for organization and are not intended to limit the disclosure in any manner. The content of any individual section may be equally applicable to all sections.
REFERENCES
The following are hereby incorporated by reference in their entireties.
[00076] A. Schulze-Bonhage, F. Sales, K. Wagner et al., "Views of patients with epilepsy on seizure prediction devices," Epilepsy & Behavior, vol. 18, no. 4, pp. 388-396, 2010.
[00077] World Health Organization. (2012, Oct.). Fact sheet on epilepsy [Online].
Available: http://wwv ho nt/mediacentre/factsheets/fs999/en/ idex.html [00078] U. R. Acharya, S. V. Sree, G. Swapna, R. J. Martis, and J. S. Suri, "Automated EEG analysis of epilepsy: a review," Knowledge-Based Systems, vol. 45, pp. 147-165.2013.
[00079] T. Alotaiby, S. A Alshebeili, T. Alshawi, 1. Ahmad, F. E. Abd El-Samie, "EEG Seizure Detection and Prediction Algorithms: A Survey," EURASIP Journal of Advances in Signal Processing (2014), DOT:10.1186/1687-6180-2014-183.
[00080] K. Gadhoumi, J. Lina, F. Monnann, J. Gotman, "Seizure Prediction for Therapeutic Devices: A Review," Journal of Neuroscience Methods, vol. 260, pp. 270282, 2015.
[00081] D. Cohen, "Magnetoencephalography, evidence of magnetic fields produced by alpha-rhythm currents," Science, 161, pp. 784-786, 1968.
[00082] M. S Ham-al-ainen, Hari, R., Ilmoniemi, R. J., Knuutila, J., Lounasman, 0. V., "Magnetoencephalography theory, instrumentation, and applications to noninvasive studies of the working human brain," Rev. Mod. Phys., vol. 65, pp. 413-497, 1993.
[00083] D. Cohen, "Magnetoencephalography: Detection of the brain's electrical activity with a superconducting magnetometer," Science, vol. 175, pp. 664-666, 1972.
[00084] X. De 'IT age, E. Carrette, B. Legros et al., "Clinical added value of magnetic source imaging in the presurgical evaluation of refractory focal epilepsy," Journal of Neurology, Neurosurgery & Psychiatry, vol. 83, no. 4, pp. 417-423, 2012.
[00085] S. Kharkar and R. Knowlton, "Magnetoencephalography in the pre-surgical evaluation of epilepsy," Epilepsy & Behavior, vol. 46, pp. 19-26, 2015.
[00086] P. Ossenblok, J. C. Dc Munck, A. Colon, W. Drolsbach, and P. Boon, "Magnetoencephalography is more successful for screening and localizing frontal lobe epilepsy than electroencephalography," Epilepsia, vol. 48, no. 11, pp. 2139-2149, 2007.
[00087] NTST/SEMATECH (2016, Mar.). e-Handbook of Statistical Methods [Online]. Available: http://www.itl.nist.govidiv898/handbook/ [00088] B.E. Boser, I.M. Guyon and V.N. Vapnik, "A training algorithm for optimal margin classifiers," Proc. 5th ACM Workshop on Computational Learning Theory (COLT), ACM Press, pp. 144-152, July 1992.
[00089] V. Vapnik, The Nature of Statistical Learning Theory, 2nd Ed., Springer, 2000.
[00090] L. Wang, Support Vector Machines: Theory and Applications, Springer Berlin Heidelberg, 2005.
[00091] A. M. Chan, E. Halgren, K. Mar nkovic, S. S. Cash. "Decoding word and category-specific spatiotemporal representations from MEG and EEG," Neurolmage, vol. 54, no. 4, pp. 3028-3039, 2011 [00092] N. F. Ince, F. Goksu, G. Pellizzer, A. Tewfik, and M. Stephane, "Selection of spectro-temporal patterns in multichannel MEG with support vector machines for schizophrenia classification." in EMBS, pp. 3554-3557, 2008.
[00093] M. T. Khalid, S. A. Aldosari, et al.,1-MEG data classification for healthy and epileptic subjects using linear discriminant analysis," IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), pp. 360-363, 2015.
[00094] F. Brig°, "Lambda waves," American J. El ectron eurodi agn osti c Technology, vol. 51, No. 2, pp. 105-13, 2011.
[000951 G. Giannakakis, V. Sakkalis, M. Pediaditis, and M. Tsiknakis, "Methods for seizure detection and prediction: An overview," Neuromethods. DOI 10.1007/7957 2014 68.
Claims (19)
- CLAIMSA computer-implemented method for in vitro detecting a brain disorder, the method comprising: (a) developing a first training set for a support vector machine (SVM) by: collecting a first Magnetoencephalography (MEG) data set of the different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the brain disorder; (b) developing a second training set for the SVM by: collecting a second MEG data set of different brain regions from that represent a healthy brain: segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (c) training the SVM to classify the state of an MEG data set based on differentiating between the first and second training sets; and (d) screening a set of MEG data by the SVM by: collecting an MEG data set of the different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets with the brain disorder.
- 2. The computer-implemented method of claim I, wherein the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism_ or facial pain.
- The computer-implemented method of claim 2, wherein the brain disorder is epilepsies.
- 4. The computer-implemented method of claim I, wherein the brain regions comprise the Left Temporal (LT), Right Temporal (KT), Left Frontal (LE), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
- The computer-implemented method of claim 1, wherein the statistical features are selected from the group comprising: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
- 6. The computer-implemented method of claim 1, wherein steps (a)-(c) constitute a training phase for the SVM.
- 7. A non-transitory computer readable medium storing specific computer-executable instructions for detecting a brain disorder that, when executed by a processor, cause a computer system to automatically at least: (a) develop a first training set for SVM by: collecting a first MEG data set of different brain regions that represent the brain disorder; segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistic& features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the brain disorder; (b) develop a second training set for the SVM by: collecting a second MEG data set of the different brain regions that represent a healthy brain; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (c) train the support vector machine to classify the state of an MEG data set based on differentiating between the first and second training sets; and (d) screen a set of MEG data by the SVM by: collecting an MEG data set of the different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets with the brain disorder.
- 8. The non-transitory computer readable medium of claim 7, wherein the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjegren's syndrome, chronic alcoholism, or facial pain.
- 9. The non-transitory computer readable medium of claim 8, wherein the brain disorder s epilepsies.
- 10. The non-transitory computer readable medium of claim 7, wherein the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
- 11. The non-transitory computer readable medium of claim 7, wherein the statistical features are selected from the group comprising: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
- 12. The non-transitory computer readable medium of cla in 7. where n steps (a)-(c) constitute a training phase for the SVM.
- 13. The non-transitory computer readable medium of claim 7, wherein the non-overlapping epochs are 1 minute epochs.
- 14. A system for detecting a brain disorder, the system comprising: (a) a computer sewer that stores MEG data sets for a plurality of subjects including subjects with the brain disorder and healthy subjects; (b) One or more computer storage media having computer-usable instructions that, when used by one or more computing devices, cause the one or more computing devices to perform a method for brain disorder risk assessment, the method comprising: (i) developing a first training set for a support vector machine (SVM) by: collecting a first MEG data set of different brain regions that represent the brain disorder: segmenting the first MEG data set of the different brain regions into non-overlapping epochs; concatenating the first MEG data set of each segment of a brain region; extracting a set of statistical features from the concatenated first MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated first MEG data set as related to the subjects of the brain disorder; (ii) developing a second training set for the SVM by: collecting a second MEG data set of the different brain regions that represent healthy subjects; segmenting the second MEG data set of the different brain regions into non-overlapping epochs; concatenating the second MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated second MEG data set of each brain region; and classifying the extracted set of statistical features from the concatenated second MEG data set as healthy; (iii) training the support vector machine to classify the state of an MEG data set based on differentiating between the first and second training sets; and (iv) screening a set of MEG data by the SVM by: collecting an MEG data set of different brain regions; segmenting the MEG data set of the different brain regions into non-overlapping epochs; concatenating the MEG data set of each segment of a brain region; extracting the set of statistical features from the concatenated MEG data set of each brain region; classifying the extracted set of statistical features from the concatenated MEG data set as representing a healthy brain or the brain disorder; and identifying MEG data sets relate to subjects with the brain disorder.
- 15, The computer-implemented method of claim 14, wherein the brain disorder is epilepsies, multiple sclerosis, Alzheimer's disease, schizophrenia, Sjogren's syndrome, chronic alcoholism, or facial pain.
- 16. The system of claim 15, wherein the brain disorder s epilepsies.
- 17. The system of claim 14, wherein the brain regions comprise the Left Temporal (LT), Right Temporal (RT), Left Frontal (LF), Right Frontal (RF), Left Parietal (LP), Right Parietal (RP), Left Occipital (LO), and Right Occipital (RO) areas of the head.
- 18. The system of claim 14, wherein the statistical features are selected from the group comprising: maximum value, minimum value, mean, standard deviation, median, interquartile range, kurtosis, and skewness.
- 19. The system of claim 14, wherein steps (i)-(iii) constitute a training phase for the SVM.
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
SA118400193 | 2018-11-13 |
Publications (3)
Publication Number | Publication Date |
---|---|
GB201900786D0 GB201900786D0 (en) | 2019-03-06 |
GB2578929A true GB2578929A (en) | 2020-06-03 |
GB2578929B GB2578929B (en) | 2021-07-07 |
Family
ID=65528307
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
GB1900786.3A Active GB2578929B (en) | 2018-11-13 | 2019-01-21 | Automatic brain disorder tendency screening using statistical features of MEG data |
Country Status (1)
Country | Link |
---|---|
GB (1) | GB2578929B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111956221A (en) * | 2020-09-07 | 2020-11-20 | 南京医科大学 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170281071A1 (en) * | 2016-03-30 | 2017-10-05 | Brain F.I.T. Imaging, LLC | Methods and magnetic imaging devices to inventory human brain cortical function |
-
2019
- 2019-01-21 GB GB1900786.3A patent/GB2578929B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170281071A1 (en) * | 2016-03-30 | 2017-10-05 | Brain F.I.T. Imaging, LLC | Methods and magnetic imaging devices to inventory human brain cortical function |
Non-Patent Citations (1)
Title |
---|
30th Annual International IEEE EMBS Conference, August 20-24 2008, Vancouver, Ince N. F. et al., "Selection of Spectro-Temporal Patterns in Multichannel MEG with Support Vector Machines for Schizophrenia Classification", pages 3554-3557 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111956221A (en) * | 2020-09-07 | 2020-11-20 | 南京医科大学 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
CN111956221B (en) * | 2020-09-07 | 2022-06-07 | 南京医科大学 | Temporal lobe epilepsy classification method based on wavelet scattering factor and LSTM neural network model |
Also Published As
Publication number | Publication date |
---|---|
GB2578929B (en) | 2021-07-07 |
GB201900786D0 (en) | 2019-03-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Jahmunah et al. | Automated detection of schizophrenia using nonlinear signal processing methods | |
Yao et al. | A robust deep learning approach for automatic classification of seizures against non-seizures | |
Kusmakar et al. | Automated detection of convulsive seizures using a wearable accelerometer device | |
Mahato et al. | Electroencephalogram (EEG) signal analysis for diagnosis of major depressive disorder (MDD): a review | |
US9101276B2 (en) | Analysis of brain patterns using temporal measures | |
US11006848B2 (en) | System and method to diagnose and predict different systemic disorders and mental states | |
Albert et al. | Automatic EEG processing for the early diagnosis of traumatic brain injury | |
Dauwels et al. | EEG synchrony analysis for early diagnosis of Alzheimer's disease: a study with several synchrony measures and EEG data sets | |
Buettner et al. | Machine learning based diagnosis of diseases using the unfolded EEG spectra: Towards an intelligent software sensor | |
CN110063732A (en) | For schizophrenia early detection and Risk Forecast System | |
Xu et al. | A graph Gaussian embedding method for predicting Alzheimer's disease progression with MEG brain networks | |
Puri et al. | Automatic detection of Alzheimer’s disease from EEG signals using low-complexity orthogonal wavelet filter banks | |
Yuan et al. | Kernel collaborative representation-based automatic seizure detection in intracranial EEG | |
WO2022135449A1 (en) | Interictal epileptiform discharge activity detection apparatus and method for epileptic patient | |
Deng et al. | SparNet: a convolutional neural network for EEG space-frequency feature learning and depression discrimination | |
Nour et al. | Diagnosis and classification of Parkinson's disease using ensemble learning and 1D-PDCovNN | |
Grover et al. | Schizo-Net: A novel Schizophrenia Diagnosis framework using late fusion multimodal deep learning on Electroencephalogram-based Brain connectivity indices | |
Susan Philip et al. | A systematic review on artifact removal and classification techniques for enhanced meg-based bci systems | |
GB2578929A (en) | Automatic brain disorder tendency screening using statistical features of MEG data | |
Kulkarni | Support vector machine based alzheimer’s disease diagnosis using synchrony features | |
Aljalal et al. | Mild Cognitive Impairment Detection from EEG Signals Using Combination of EMD Decomposition and Machine Learning | |
Boppana et al. | Deep Learning Approach for an early stage detection of Neurodevelopmental Disorders | |
Ranjan et al. | A machine learning framework for automatic diagnosis of schizophrenia using EEG signals | |
Anchana et al. | Comparative study of detection of ADHD using EEG signals | |
Zhang et al. | Comparison of Neural Networks' Performance in Early Screening of Autism Spectrum Disorders Under Two MRI Principles |