WO2024040423A1 - 基于脑电图的急性意识障碍预后评估方法、装置、介质 - Google Patents

基于脑电图的急性意识障碍预后评估方法、装置、介质 Download PDF

Info

Publication number
WO2024040423A1
WO2024040423A1 PCT/CN2022/114126 CN2022114126W WO2024040423A1 WO 2024040423 A1 WO2024040423 A1 WO 2024040423A1 CN 2022114126 W CN2022114126 W CN 2022114126W WO 2024040423 A1 WO2024040423 A1 WO 2024040423A1
Authority
WO
WIPO (PCT)
Prior art keywords
waveform
electroencephalogram
characteristic
evaluation
acute
Prior art date
Application number
PCT/CN2022/114126
Other languages
English (en)
French (fr)
Inventor
王星
Original Assignee
北京太阳电子科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 北京太阳电子科技有限公司 filed Critical 北京太阳电子科技有限公司
Priority to PCT/CN2022/114126 priority Critical patent/WO2024040423A1/zh
Publication of WO2024040423A1 publication Critical patent/WO2024040423A1/zh

Links

Images

Classifications

    • 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
    • A61B5/24—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]
    • A61B5/372—Analysis of electroencephalograms

Definitions

  • the present disclosure relates to the technical field of electroencephalogram applications, and in particular, to an electroencephalogram-based prognosis assessment method, device, and medium for acute consciousness disorder.
  • Acute disorder of consciousness is a neurological disease caused by damage to the central nervous system leading to impairment of individual perception. It may be caused by a variety of causes such as acute ischemia and hypoxia, cerebrovascular disease, intracranial inflammation, poisoning, and metabolic diseases.
  • the prognostic assessment of acute disturbance of consciousness can assist doctors in adjusting treatment measures in a timely manner and promote the patient's recovery of consciousness.
  • Electroencephalogram is a bioelectric signal recorded through electrodes placed in the cerebral cortex or intracranium. Due to its characteristics of non-invasive, continuous, real-time, dynamic, cheap, and highly correlated with the state of consciousness, it is widely used in It plays a very important role in the prognosis of disorders of consciousness.
  • EEG Electroencephalogram
  • the present disclosure provides an electroencephalogram-based prognosis assessment method, device, and medium for acute consciousness disorder.
  • an electroencephalogram-based prognostic assessment method for acute consciousness disorder including: obtaining the patient's electroencephalogram data;
  • the characteristic waveforms are classified and corrected to form a waveform set.
  • the waveform set includes at least one characteristic waveform group.
  • Each of the characteristic waveform groups includes a plurality of related characteristic waveforms belonging to the same kind of characteristic waveform. waveform segment;
  • the characteristic waveforms include alpha rhythm, theta rhythm, spindles, delta waves, burst suppression, electrostatic rest, epileptiform discharges, triphasic waves, and/or GPD.
  • classification processing and error correction processing are performed on the characteristic waveforms to form a waveform set, including:
  • the pre-stored multiple waveform evaluation rules include scale evaluation rules and prediction network model evaluation rules.
  • the scale evaluation rules include at least one of Synek, Yang, and Hofmeijer; and/or,
  • the prediction network model evaluation rules include multiple sub-evaluation network models.
  • the sub-evaluation network models are formed based on electroencephalogram parameter training of patients in a coma state. Different sub-evaluation network models correspond to the duration of the patient's coma state. different.
  • the waveform set is evaluated based on a variety of pre-stored waveform evaluation rules, including:
  • the prognostic evaluation results corresponding to the waveform set include prognostic evaluation, the position of the characteristic waveform in the electroencephalogram data, and the evaluation score, wherein the prognostic evaluation Including good prognosis and poor prognosis.
  • an electroencephalogram-based prognosis assessment device for acute consciousness disorder includes:
  • a receiving module configured to obtain electroencephalogram data of the patient
  • An extraction module configured to identify the electroencephalogram data and extract characteristic waveforms in the electroencephalogram data
  • a clustering module configured to classify and correct the characteristic waveforms to form a waveform set.
  • the waveform set includes at least one characteristic waveform group, and each of the characteristic waveform groups includes multiple characteristic waveform groups belonging to the same type. Multiple related waveform segments of the characteristic waveform;
  • An evaluation module configured to evaluate the waveform set based on a variety of pre-stored waveform evaluation rules
  • An output module is configured to output the prognosis evaluation results corresponding to the waveform set under each of the waveform evaluation rules.
  • an electroencephalogram-based prognosis assessment device for acute consciousness disorder includes:
  • Memory used to store instructions executable by the processor
  • the processor is configured to execute the electroencephalogram-based prognosis assessment method for acute consciousness disorder provided by the first aspect of the present disclosure.
  • a non-transitory computer-readable storage medium when instructions in the storage medium are executed by a processor of an electroencephalogram-based acute consciousness disorder prognostic assessment device when the processor is enabled to execute the electroencephalogram-based prognosis assessment method for acute consciousness disorder provided by the embodiment provided in the first aspect of the present disclosure.
  • a waveform set is formed. Based on a variety of pre-stored waveform evaluation rules, the waveforms in the waveform set are evaluated, and the prognosis assessment corresponding to the waveform set under each waveform evaluation rule is output. result. It not only simplifies the operation process and realizes automated grading, but also significantly improves the reliability, interpretability and accuracy of prognostic assessment of patients with acute disorders of consciousness.
  • FIG. 1 is a flow chart of an electroencephalogram-based prognostic assessment method for acute consciousness disorder according to an exemplary embodiment.
  • FIG. 2 is a block diagram of an electroencephalogram-based prognosis assessment device for acute consciousness disorder according to an exemplary embodiment.
  • FIG. 3 is a block diagram of an electroencephalogram-based prognosis assessment device for acute consciousness disorder according to an exemplary embodiment.
  • Exemplary embodiments of the present disclosure provide an electroencephalogram-based prognostic assessment method for acute consciousness disorders.
  • the electroencephalogram-based prognostic assessment method for acute consciousness disorder shown in this disclosure includes:
  • the waveform set includes at least one characteristic waveform group, and each characteristic waveform group includes multiple related waveform segments belonging to the same characteristic waveform;
  • S105 Output the prognostic evaluation results corresponding to the waveform set under each waveform evaluation rule.
  • the EEG signals of patients who have been diagnosed with acute disorders of consciousness or who are prone to acute disorders of consciousness are collected.
  • the collection method can be, for example, by attaching electrodes to the patient's head to collect the patient's EEG data.
  • Specialized EEG acquisition equipment can be used for collection.
  • the number of electrodes can be selected based on the patient's condition and the doctor's experience. For example, if the patient's symptoms are more severe, the number of electrodes can be increased; for another example, if the patient's brain is partially traumatized, the number of electrodes can be increased after the patient's symptoms are severe. A smaller number of electrodes are attached to the location of the wound, and no electrodes are attached to the other locations.
  • the positions of the electrodes can be placed according to the international standard lead 10-20 electrode system (that is, the international standard 10-20 for electroencephalogram electrode positions), and the number of electrodes can be appropriately reduced according to clinical needs, such as setting 18 Electrodes are used to obtain the patient's electroencephalogram data as basic data for prognostic assessment.
  • the international standard lead 10-20 electrode system is a common reference standard for the EEG data collection process in this field.
  • the acquired EEG data is input to a device such as a characteristic waveform detector to extract the characteristic waveforms in the EEG data. For example, by traversing the EEG data, extracting the characteristic waveform from the EEG data. Characteristic waveforms related to acute disturbance of consciousness are extracted from the graph data. If the device used to collect EEG data also has a characteristic waveform detection function, the characteristic waveform in the EEG data can be directly output by the device that collects EEG data.
  • characteristic waveforms include alpha rhythm, theta rhythm, spindle waves, delta waves, burst suppression, electrical rest, epileptiform discharges, triphasic waves and/or GPD.
  • relevant characteristic waveforms can be selectively extracted based on the patient's state and the physician's experience. For example, alpha rhythm, triphasic waves, and burst suppression can be extracted for patients with encephalopathy. For example, alpha rhythm, theta rhythm, spindle waves, and GPD can be extracted while the patient is sleeping.
  • alpha rhythm is a rhythmic waveform with a frequency of 8-12Hz in EEG. Most of the waveforms are sinusoidal. It is easy to appear when the subject is closed eyes and relaxed. It is the most important factor for analyzing the background activity of EEG. index.
  • Theta rhythm is a rhythmic waveform with a frequency of 12-20Hz in the electroencephalogram, which occurs during normal adult sleep.
  • Spindle also known as ⁇ rhythm, is a sign of sleep entering the second stage and can continue to the third stage. It is one of the many important measurable characteristics of brain activity during sleep.
  • Delta wave is a waveform with a frequency below 4Hz in the electroencephalogram.
  • Burst-suppression is a severe electroencephalogram abnormality that manifests as burst activity of medium to high amplitude, waveforms that alternate with low voltage or electrical suppression states. It is a symptom of extensive damage to the cerebral cortex and subcortex or manifestations of inhibition. Electrical rest is a waveform in which brain electrical activity is continuously lower than 20uV. It is not affected by state changes and rarely responds to external stimuli. Clinically, epileptiform discharges include paroxysmal abnormalities such as spike waves, sharp waves, spike-slow complexes, sharp-slow complex waves, and polyspike-slow complex waves.
  • Three-phase waves are medium-high amplitude slow waves with a frequency of about 1-2 Hz.
  • the brain waves deflect three times up and down along the baseline, forming a wave with three components alternating up and down the baseline.
  • the first phase is a negative wave with lower amplitude
  • the second phase It is a prominent forward wave
  • the third phase is a negative phase slow wave with a longer duration than the second phase.
  • GPD Generalized Periodic Discharge
  • Compound waves such as generalized spike waves and sharp waves that stand out from the background appear repeatedly and stereotypedly at similar intervals. It is a seriously abnormal electroencephalogram phenomenon.
  • step S103 the characteristic waveforms extracted in step S102 are classified according to waveform categories, and multiple related waveform segments belonging to the same characteristic waveform are classified into the same characteristic waveform group. For example, if ⁇ rhythm, three-phase If there are three waveforms: alpha rhythm, triphasic wave, and burst suppression, then each waveform segment belonging to alpha rhythm, triphasic wave, and burst suppression in the EEG data is classified together to form three characteristics of ⁇ rhythm, triphasic wave, and burst suppression. Waveform group. At the same time, error correction is performed on the waveforms of each set of characteristic waveforms to eliminate irrelevant waveforms incorrectly extracted by the waveform detector.
  • the waveform detector actually extracts the theta rhythm waveform segment, which is misjudged as ⁇ rhythm during output and is classified into In the ⁇ rhythm characteristic waveform group, after error correction processing, the theta rhythm waveform segments can be eliminated, so that all the characteristic waveform groups in the characteristic waveform group contain correct corresponding waveform segments.
  • All characteristic waveform groups constitute a waveform set, and the number of characteristic waveform groups is not limited. Everything is subject to the number of extracted characteristic waveform categories.
  • step S103 the characteristic waveforms are classified and corrected to form a waveform set, including:
  • S1031 Group waveform segments belonging to the same characteristic waveform into one category as a set of characteristic waveform groups, and multiple waveform segments in each group of characteristic waveform segments are mutually related waveform segments;
  • S1032 Perform time-frequency transformation on each relevant waveform segment in each characteristic waveform group, and obtain the frequency domain information of each relevant waveform segment in each characteristic waveform group;
  • S1033 Calculate the average value of multiple frequency domain information, and use the obtained average value as the reference frequency domain information;
  • S1034 Calculate the difference between the frequency domain information and the reference frequency domain information, and remove the relevant waveform segments corresponding to the frequency domain information whose difference is greater than the preset value from the characteristic waveform group to form a waveform set.
  • step S1031 attributes all extracted ⁇ rhythms in special waveforms to the ⁇ rhythm characteristic waveform group, so that all waveform segments in the ⁇ rhythm characteristic waveform group are ⁇ rhythm.
  • step S1032 time-frequency transformation is performed on each ⁇ -rhythm waveform segment in the ⁇ -rhythm characteristic waveform group to obtain conventional frequency domain information such as frequency and amplitude of each waveform segment.
  • step S1033 synthesizes the frequency domain information of all ⁇ rhythm waveform segments in the group, obtains the average value through a clustering algorithm, and uses the average value as the reference frequency domain information of the ⁇ rhythm characteristic waveform group for error correction processing in step S1034.
  • Step S1034 calculates the difference between the frequency domain information of each ⁇ rhythm waveform segment and the reference frequency domain information, sorts them according to the size of the difference, and eliminates the ⁇ rhythm waveform segment corresponding to the frequency domain information whose difference is greater than the preset value.
  • the ⁇ rhythm waveform segments that meet the preset difference requirements are retained and determined as the final ⁇ rhythm waveform group, where the preset value of the difference can be set individually according to the properties of each characteristic waveform.
  • each extracted characteristic waveform segment such as theta rhythm, spindle wave, GPD, etc. is subjected to the same classification and error correction processing to form multiple groups of characteristic waveform groups, and all the characteristic waveform groups constitute a waveform set.
  • the pre-stored waveform rules are all mature grading methods at this stage, including manual grading methods and semi-automatic grading methods. These mature grading methods are added as evaluation rules to the prognosis evaluation method, among which, according to clinical needs, The evaluation rules are selectively used, and multiple pre-stored evaluation rules are used to evaluate the waveform collection multiple times, making the evaluation results of the waveform collection interpretable and reliable.
  • the various pre-stored waveform evaluation rules include scale evaluation rules and prediction network model evaluation rules.
  • the scale evaluation rule is the current manual grading method.
  • the sensitivity and specificity of the related grading scheme are not high, requiring doctors to have rich experience in chart reading and electroencephalogram expertise.
  • the prediction network model evaluation rule is the current semi-automatic grading method. This method has a single feature dimension and poor interpretability, and needs clinical data verification support. Adding existing manual grading methods and semi-automatic grading methods as pre-stored evaluation rules to the prognostic evaluation method simplifies the operation process, makes the evaluation results output by automated grading more theoretically supported, and improves interpretability and reliability.
  • the scale evaluation rules include at least one of Synek, Yang, and Hofmeijer.
  • the prediction network model evaluation rules include multiple sub-evaluation network models.
  • the sub-evaluation network models are formed based on the electroencephalogram parameter training of the patient in a coma. Different sub-evaluation network models correspond to different durations of the patient's coma.
  • Synek grading is an electroencephalogram-based prognostic judgment scale for adult coma patients proposed by Synek in 1988. It is the first grading system to use electroencephalogram signals as the main information. It is generally divided into Level 4, this grading method has strong theoretical and practical support.
  • the Young classification was proposed by Young in 1997 based on the Synek classification. As a grading standard for severe brain functional injury, its classification is more widely used.
  • the Hofmeijer classification is for patients after cardiopulmonary resuscitation and is generally divided into three levels.
  • the prediction network model uses a convolutional neural network with a VGG architecture, and is trained to form sub-assessments of different durations by collecting brain wave parameters of different durations when the patient is in a coma, such as 12 hours and 24 hours of coma. network model.
  • the waveform collection of the patient's electroencephalogram is used as input data and input into the prediction network model for evaluation.
  • the corresponding sub-evaluation network model needs to be used as the evaluation rule.
  • the patient is already in a coma. If the state is 12 hours, it corresponds to the sub-evaluation network model using the coma for 12 hours.
  • the waveform collection of different durations when the patient is in a coma state can be used as EEG data for subsequent neural network training.
  • step S104 evaluates the waveform set based on a variety of pre-stored waveform evaluation rules, including:
  • S1041 Match the reference waveform included in the waveform evaluation rule with multiple relevant waveform segments in each characteristic waveform group in the waveform set, and obtain the first matching degree between each relevant waveform segment and its corresponding reference waveform;
  • S1042 According to the first matching degree, obtain the second matching degree of each characteristic waveform group;
  • the various waveform evaluation rules prestored in step S1041 all include the reference waveform of each characteristic waveform and its related information.
  • the reference waveform is used as the reference standard for evaluation, and each relevant waveform segment in each characteristic waveform group is of the same category.
  • the reference waveform is matched, and the matching degree of each relevant waveform segment is recorded as the first matching degree.
  • step S1042 after each relevant waveform segment in each characteristic waveform group is matched, all first matching degrees are combined to obtain the second matching degree of each characteristic waveform group.
  • Step S1043 combines the second matching degree of each characteristic waveform group to obtain the evaluation score of the waveform set under different evaluation rules.
  • step S105 the prognostic evaluation results of the waveform set evaluated by each grading method are output, and the combined output results carry prediction evidence, which improves the reliability of the prognostic evaluation.
  • the prognostic evaluation results corresponding to the waveform set include prognostic evaluation, the position of the characteristic waveform in the electroencephalogram data, and the evaluation score, where the prognostic evaluation includes good prognosis and poor prognosis.
  • the prognostic evaluation results include evaluation scores and also provide the location and location of relevant characteristic waveforms in the EEG data.
  • the original image provides an index for doctors to view the results, making it easier for doctors to lock the information in the EEG and find the cause of the disease when analyzing the patient's condition.
  • the prognosis evaluation has two criteria: good prognosis and poor prognosis, to provide doctors with the most concise evaluation results and facilitate doctors to judge different levels of treatment for patients in the future. For example, if the prognosis is good, a more aggressive treatment plan can be adopted to help the patient recover as quickly as possible. If the prognosis evaluation is poor, cheaper drugs can be used during treatment and conservative treatment can be used to help patients reduce their financial burden.
  • Exemplary embodiments of the present disclosure provide an electroencephalogram-based prognosis assessment device for acute consciousness disorder.
  • FIG. 2 a block diagram of the electroencephalogram-based prognosis assessment device for acute consciousness disorder shown in the present disclosure.
  • the block diagram includes: receiving module 21, extraction module 22, clustering module 23, evaluation module 24 and output module 25.
  • the receiving module 21 is configured to obtain the patient's electroencephalogram data
  • the extraction module 22 is configured to identify EEG data and extract characteristic waveforms in the EEG data
  • the clustering module 23 is configured to classify and correct the characteristic waveforms to form a waveform set.
  • the waveform set includes at least one characteristic waveform group, and each characteristic waveform group includes a plurality of related characteristic waveforms belonging to the same characteristic waveform. waveform segment;
  • the evaluation module 24 is configured to evaluate the waveform set based on a variety of pre-stored waveform evaluation rules
  • the output module 25 is configured to output the prognostic evaluation results corresponding to the waveform set under each waveform evaluation rule.
  • the characteristic waveforms are classified and corrected to form a waveform set, and the clustering module 23 is configured as:
  • evaluation module 24 is configured to:
  • FIG. 3 is a block diagram of an electroencephalogram-based prognosis assessment device 1000 for acute consciousness disorder according to an exemplary embodiment.
  • the device 1000 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, or the like.
  • the device 1000 may include one or more of the following components: a processing component 1002, a memory 1004, a power component 1006, a multimedia component 1008, an audio component 1010, an input/output (I/O) interface 1012, a sensor component 1014, and communications component 1016.
  • Processing component 1002 generally controls the overall operations of device 1000, such as operations associated with display, phone calls, data communications, camera operations, and recording operations.
  • the processing component 1002 may include one or more processors 1020 to execute instructions to complete all or part of the steps of the above method.
  • processing component 1002 may include one or more modules that facilitate interaction between processing component 1002 and other components.
  • processing component 1002 may include a multimedia module to facilitate interaction between multimedia component 1008 and processing component 1002.
  • Memory 1004 is configured to store various types of data to support operations at device 1000 . Examples of such data include instructions for any application or method operating on device 1000, contact data, phonebook data, messages, pictures, videos, etc.
  • Memory 1004 may be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EEPROM), Programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EEPROM erasable programmable read-only memory
  • EPROM Programmable read-only memory
  • PROM programmable read-only memory
  • ROM read-only memory
  • magnetic memory flash memory, magnetic or optical disk.
  • Power component 1006 provides power to various components of device 1000.
  • Power components 1006 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to device 1000 .
  • Multimedia component 1008 includes a screen that provides an output interface between device 1000 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user.
  • the touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. A touch sensor can not only sense the boundaries of a touch or swipe action, but also detect the duration and pressure associated with the touch or swipe action.
  • multimedia component 1008 includes a front-facing camera and/or a rear-facing camera.
  • the front camera and/or the rear camera may receive external multimedia data.
  • Each front-facing camera and rear-facing camera can be a fixed optical lens system or have a focal length and optical zoom capabilities.
  • Audio component 1010 is configured to output and/or input audio signals.
  • audio component 1010 includes a microphone (MIC) configured to receive external audio signals when device 1000 is in operating modes, such as call mode, recording mode, and speech recognition mode. The received audio signals may be further stored in memory 1004 or sent via communication component 1016 .
  • audio component 1010 also includes a speaker for outputting audio signals.
  • the I/O interface 1012 provides an interface between the processing component 1002 and a peripheral interface module.
  • the peripheral interface module may be a keyboard, a click wheel, a button, etc. These buttons may include, but are not limited to: Home button, Volume buttons, Start button, and Lock button.
  • Sensor component 1014 includes one or more sensors for providing various aspects of status assessment for device 1000 .
  • the sensor component 1014 can detect the open/closed state of the device 1000, the relative positioning of components, such as the display and keypad of the device 1000, the sensor component 1014 can also detect the position change of the device 1000 or a component of the device 1000, the user The presence or absence of contact with device 1000, device 1000 orientation or acceleration/deceleration and temperature changes of device 1000.
  • Sensor assembly 1014 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact.
  • Sensor assembly 1014 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications.
  • the sensor component 1014 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • Communication component 1016 is configured to facilitate wired or wireless communication between apparatus 1000 and other devices.
  • Device 1000 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof.
  • the communication component 1016 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • communications component 1016 also includes a near field communications (NFC) module to facilitate short-range communications.
  • NFC near field communications
  • the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • apparatus 1000 may be configured by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable Gate array (FPGA), controller, microcontroller, microprocessor or other electronic components are implemented for executing the above method.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable Gate array
  • controller microcontroller, microprocessor or other electronic components are implemented for executing the above method.
  • non-transitory computer-readable storage medium including instructions, such as a memory 1004 including instructions, which can be executed by the processor 1020 of the device 1000 to complete the above method is also provided.
  • non-transitory computer-readable storage media may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
  • a non-transitory computer-readable storage medium that, when instructions in the storage medium are executed by a processor of a processing device of an electronic device, enables the processing device of the electronic device to execute what is provided by the exemplary embodiments of the present disclosure based on the brain. Electrogram-based prognostic assessment of acute disorders of consciousness.
  • the embodiment of the present disclosure extracts relevant characteristic waveforms of acute consciousness disorder in electroencephalogram to form a waveform set, converts existing scales and prediction network models into evaluation rules, analyzes the waveforms in the waveform set, and outputs prediction evidence
  • the prognostic assessment results can greatly simplify the operation process, realize automated grading, and significantly improve the reliability, interpretability and accuracy of prognostic assessment of patients with acute disorders of consciousness.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Psychology (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Engineering & Computer Science (AREA)
  • Psychiatry (AREA)
  • Physics & Mathematics (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

本公开是关于一种基于脑电图的急性意识障碍预后评估方法、装置及介质，基于脑电图的急性意识障碍预后评估方法包括：获取患者的脑电图数据；对脑电图数据进行识别，提取脑电图数据中的特征波形；对特征波形进行分类处理和纠错处理，形成波形集合，波形集合包括至少一组特征波形组，每组特征波形组包括多个属于同一种特征波形的多个相关波形段；基于预存的多种波形评估规则，对波形集合进行评估；输出每种波形评估规则下，与波形集合对应的预后评估结果。采用本公开中基于脑电图的急性意识障碍预后评估方法，可以简化操作过程，实现自动化分级，显著提升急性意识障碍患者预后评估的可靠性、可解释性和准确性。

Description

基于脑电图的急性意识障碍预后评估方法、装置、介质 技术领域
本公开涉及脑电图应用技术领域，尤其涉及一种基于脑电图的急性意识障碍预后评估方法、装置、介质。
背景技术
急性意识障碍是由中枢神经系统受损导致个体感知能力出现障碍的一种神经系统疾病，可能由于急性缺血缺氧、脑血管病、颅内炎症、中毒及代谢性疾病等多种病因引起。急性意识障碍的预后评估能够辅助医师及时调整治疗措施，促进患者的意识恢复。脑电图(electroencephalogram，EEG)是一种通过安放在大脑皮质或颅内的电极所记录到的生物电信号，由于其无创、连续、实时、动态、廉价、与意识状态相关性高等特点，在意识障碍预后判断中起到了非常重要的作用。但临床中采用EEG进行预后评估需要依靠专业医生对长程脑电记录进行判读，工作量巨大，并且判读结论一致性差，非常依赖医生经验与专业素养。
发明内容
为克服相关技术中存在的问题，本公开提供一种基于脑电图的急性意识障碍预后评估方法、装置、介质。
根据本公开实施例的第一方面，提供了一种基于脑电图的急性意识障碍预后评估方法，包括：获取患者的脑电图数据；
对所述脑电图数据进行识别，提取所述脑电图数据中的特征波形；
对所述特征波形进行分类处理和纠错处理，形成波形集合，所述波形集合包括至少一组特征波形组，每组所述特征波形组包括多个属于同一种所述特征波形的多个相关波形段；
基于预存的多种波形评估规则，对所述波形集合进行评估；
输出每种所述波形评估规则下，与所述波形集合对应的预后评估结果。
在本公开一些示例性的实施例中，所述特征波形包括α节律、θ节律、纺锤波、δ波、爆发抑制、电静息、癫痫样放电、三相波和/或GPD。
在本公开一些示例性的实施例中，对所述特征波形进行分类处理和纠错处理，形成波形集合，包括：
将属于同一种所述特征波形的波形段归为一类，作为一组所述特征波形组，每组所述特征波形段中的多个波形段互为相关波形段；
对每组所述特征波形组中的每个相关波形段进行时频变换，获得每组所述特征波形组中的每个相关波形段的频域信息；
对多个所述频域信息进行平均值计算，将获得的平均值作为参考频域信息；
计算所述频域信息与所述参考频域信息之间的差值，将所述差值大于预设值的所述频域信息对应的所述相关波形段从所述特征波形组中剔除，形成所述波形集合。
在本公开一些示例性的实施例中，所述预存的多种波形评估规则包括量表评估规则和预测网络模型评估规则。
在本公开一些示例性的实施例中，所述量表评估规则包括Synek、Yang、Hofmeijer中的至少一种；和/或，
所述预测网络模型评估规则包括多个子评估网络模型，所述子评估网络模型基于患者处于昏迷状态的脑电图参数训练形成，不同的所述子评估网络模型对应的患者处于昏迷状态的持续时长不同。
在本公开一些示例性的实施例中，基于预存的多种波形评估规则，对所述波形集合进行评估，包括：
基于所述波形评估规则包含的参考波形与所述波形集合中的每组所述特征波形组中的多个相关波形段进行匹配，获得每个相关波形段和与其对应的所述参考波形之间的第一匹配度；
根据所述第一匹配度，获得每组所述特征波形组的第二匹配度；
根据所述第二匹配度，获得所述波形集合的评估得分。
在本公开一些示例性的实施例中，所述波形集合对应的预后评估结果包括预后评价、所述特征波形在所述脑电图数据中的位置和所述评估得分，其中，所述预后评价包括预后状况好和预后状况差。
根据本公开实施例的第二方面，提供了一种基于脑电图的急性意识障碍预后评估装置，所述基于脑电图的急性意识障碍预后评估装置包括：
接收模块，被配置为获取患者的脑电图数据；
提取模块，被配置为对所述脑电图数据进行识别，提取所述脑电图数据中的特征波形；
聚类模块，被配置为对所述特征波形进行分类处理和纠错处理，形成波形集合，所述波形集合包括至少一组特征波形组，每组所述特征波形组包括多个属于同一种所述特征波形的多个相关波形段；
评估模块，被配置为基于预存的多种波形评估规则，对所述波形集合进行评估；
输出模块，被配置为输出每种所述波形评估规则下，与所述波形集合对应的预后评估结果。
根据本公开实施例的第三方面，提供了一种基于脑电图的急性意识障碍预后评估装置，所述基于脑电图的急性意识障碍预后评估装置包括：
处理器；
用于存储处理器可执行指令的存储器；
其中，所述处理器被配置为执行本公开第一方面提供的基于脑电图的急性意识障碍预后评估方法。
根据本公开示例性的实施例的第四方面，提供了一种非临时性计算机可读存储介质，当所述存储介质中的指令由基于脑电图的急性意识障碍预后评估装置的处理器执行时，使得处理器能够执行本公开第一方面提供的实施例所提供的基于脑电图的急性意识障碍预后评估方法。
本公开的实施例提供的技术方案可以包括以下有益效果：
通过提取脑电图中急性意识障碍的特征波形，形成波形集合，基于预存的多种波形评估规则，对波形集合中的波形进行评估，输出每种波形评估规则下，与波形集合对应的预后评估结果。不仅简化了操作过程，实现自动化分级，还显著提升了急性意识障碍患者预后评估的可靠性、可解释性和准确性。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本公开。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本发明的实施例，并与说明书一起用于解释本发明的原理。
图1是根据一示例性实施例示出的基于脑电图的急性意识障碍预后评估方法的流程图。
图2是根据一示例性实施例示出的基于脑电图的急性意识障碍预后评估装置的框图。
图3是根据一示例性实施例示出的基于脑电图的急性意识障碍预后评估装置的框图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时， 除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本发明相一致的所有实施方式。相反，它们仅是与如所附权利要求书中所详述的、本发明的一些方面相一致的装置和方法的例子。
本公开示例性的实施例提供了一种基于脑电图的急性意识障碍预后评估方法。如图1所示，本公开中示出的基于脑电图的急性意识障碍预后评估方法，包括：
S101：获取患者的脑电图数据；
S102：对脑电图数据进行识别，提取脑电图数据中的特征波形；
S103：对特征波形进行分类处理和纠错处理，形成波形集合，波形集合包括至少一组特征波形组，每组特征波形组包括多个属于同一种特征波形的多个相关波形段；
S104：基于预存的多种波形评估规则，对波形集合进行评估；
S105：输出每种波形评估规则下，与波形集合对应的预后评估结果。
其中，在步骤S101中，采集已被诊断为急性意识障碍或者有急性意识障碍倾向患者的脑电信号，采集方式比如可以通过在患者的头部贴附电极的方式采集患者的脑电图数据，可以采用专门的脑电图采集设备进行采集。在贴附电极时，电极的数量可以根据患者的状态以及医师的经验进行选择，比如，患者症状比较严重，则可以增加电极数量；再比如，患者的脑部局部位置受到创伤，则可以在受到创伤的位置贴附数量较少的电极，其余位置不用贴附。在一个示例中，电极的位置可以按照国际标准导联10-20电极系统(也即脑电图电极位置国际标准10-20)安放，电极的数量可以根据临床需要做适当减少，比如设置18个电极，以此获取患者的脑电图数据作为预后评估的基础数据。其中，国际标准导联10-20电极系统是本领域中通用的一个脑电图数据采集过程的参照标准。
在步骤S102中，将获取的脑电图数据输入至比如特征波形检测器等设备中，以对脑电图数据中的特征波形进行提取，比如，采用遍历脑电图数据的方式，从脑电图数据中提取出与急性意识障碍相关的特征波形。如果用于采集脑电图数据的设备同时具有特征波形检测功能，则可以直接由采集脑电图数据的设备输出脑电图数据中的特征波形。
其中，特征波形包括α节律、θ节律、纺锤波、δ波、爆发抑制、电静息、癫痫样放电、三相波和/或GPD。在对脑电图数据中的特征波形进行提取时，可以根据患者的状态以及医师的经验选择性提取相关特征波形，比如，对患有脑病的患者提取α节律、三相波、爆发抑制，再比如，患者处于睡眠状态下提取α节律、θ节律、纺锤波、GPD。
其中，α节律是脑电图中频率处于8-12Hz的节律形波形，波形大多为正弦样，当被采集者处于闭眼且精神放松状态下容易出现，是分析脑电图背景活动最重要的指标。θ节律是脑 电图中频率处于12-20Hz的节律形波形，在正常成人睡眠时出现。纺锤波(sleep spindle)又称为σ节律，是睡眠进入二期的标志，可以延续到三期，是睡眠期间大脑活动的许多重要可测量特征之一。δ波是是脑电图中频率为4Hz以下的波形，其中正常成人在清醒状态下几乎不会出现δ波，只有在睡眠状态下会出现。爆发抑制(burst-suppression)是一种严重的脑电图失常现象，表现为中-高波幅的爆发性活动，与低电压或电抑制状态交替出现的波形，是大脑皮质和皮质下广泛损伤或抑制的表现。电静息是脑电活动持续低于20uV的波形,不受状态变化的影响，对外界刺激很少有反应。癫痫样放电在临床上包括棘波、尖波、棘慢复合波、尖慢复合波、多棘慢复合波等阵发性异常。三相波是频率约为1-2Hz的中高幅慢波，脑波沿基线上下有三次偏转，形成基线上下交替三个成分的波，第一相为波幅较低的负向波，第二相为突出的正向波，第三相为时限长于第二相的负相慢波，多见于多种代谢性脑病以及克雅病，缺氧、中毒性脑病等中。GPD(Generalized Periodic Discharge)，被称为广泛性周期性复合波，突出于背景的广泛性棘波、尖波等复合波以相似的间隔反复刻板出现，是一种严重异常的脑电图现象。
在步骤S103中，把在步骤S102中提取到的特征波形按照波形类别分类，将属于同一种特征波形的多个相关波形段归到同一特征波形组中，比如，若提取了α节律、三相波、爆发抑制三种波形，则分别将脑电图数据中的属于α节律、三相波、爆发抑制的每个波形段归类到一起，形成α节律、三相波、爆发抑制三个特征波形组。同时，对每组特征波形组波形进行纠错，剔除波形检测器错误提取的无关波形，比如，波形检测器实际提取的是θ节律波形段，在输出时误判为α节律，被归类到α节律的特征波形组，经过纠错处理，可以把θ节律波形段剔除，使得特征波形组中均为正确的对应波形段。所有特征波形组构成波形集合，其中特征波形组的数量不做限定，一切以提取到的特征波形类别的数量为准。
其中，步骤S103中对特征波形进行分类处理和纠错处理，形成波形集合，包括：
S1031：将属于同一种特征波形的波形段归为一类，作为一组特征波形组，每组特征波形段中的多个波形段互为相关波形段；
S1032：对每组特征波形组中的每个相关波形段进行时频变换，获得每组特征波形组中的每个相关波形段的频域信息；
S1033：对多个频域信息进行平均值计算，将获得的平均值作为参考频域信息；
S1034：计算频域信息与参考频域信息之间的差值，将差值大于预设值的频域信息对应的相关波形段从特征波形组中剔除，形成波形集合。
其中，以α节律为例，步骤S1031把提取到的所有特种波形中的α节律，归属到α节律 特征波形组中，使α节律特征波形组中的波形段均为α节律。在步骤S1032中，对α节律特征波形组中的每个α节律波形段均进行时频变换，得到每个波形段的频率、波幅等常规频域信息。步骤S1033综合该组中所有α节律波形段的频域信息，通过聚类算法得到平均值，将该平均值作为α节律特征波形组的参考频域信息，以供步骤S1034进行纠错处理。步骤S1034计算每个α节律波形段的频域信息与参考频域信息之间的差值，按照差值大小进行排序，并剔除差值大于预设值的频域信息对应的α节律波形段，保留符合预设置差值要求的α节律波形段并确定为最终的α节律波形组，其中，差值的预设值可以按照每个特征波形的性质进行单独设置。同理对θ节律、纺锤波、GPD等提取到的每个特征波形段进行相同的分类处理和纠错处理的，构成多组特征波形组，由所有特征波形组构成波形集合。通过分类处理和纠错处理，可以排除由于波形检测器的检测误差造成的错误提取非相关波形，减少对预后评估结果的影响，使结果更为准确。
步骤S104中，预存的波形规则均为现阶段成熟的分级方法，其中包括手动分级方法和半自动分级方法，将这些成熟的分级方法作为评估规则一同加入预后评估方法中，其中，可以根据临床需求从评估规则中进行选择性使用，由多种预存的评估规则对波形集合进行多次评估，使得对波形集合的评估结果具有可解释性和可靠性。
其中，预存的多种波形评估规则包括量表评估规则和预测网络模型评估规则。
量表评估规则是现阶段的手动分级方法，相关分级方案的灵敏度与特异性不高，需要医生具有丰富的读图经验和脑电图专业知识储备，预测网络模型评估规则是现阶段的半自动分级方法，其应用的特征维度单一且可解释性差，需要临床数据验证支持。将现有的手动分级方法和半自动分级方法作为预存的评估规则共同加入预后评估方法中，简化了操作过程，可以使自动化分级输出的评估结果更具有理论支持，提高可解释性和可靠性。
其中，量表评估规则包括Synek、Yang、Hofmeijer中的至少一种。预测网络模型评估规则包括多个子评估网络模型，子评估网络模型基于患者处于昏迷状态的脑电图参数训练形成，不同的子评估网络模型对应的患者处于昏迷状态的持续时长不同。
在手动分级方法中，Synek分级是由Synek在1988年提出的一种基于脑电图的成年昏迷患者预后判断量表，是第一个用脑电图信号作为主要信息的分级体系，总体分为四级，该分级方法具有很强的理论及实践支持。Young分级是由Young 1997年在Synek分级的基础上提出，作为一种重症脑功能损伤分级标准，其分级的应用更为广泛。Hofmeijer分级是针对心肺复苏后患者，总体分为三级。在半自动分级方法中，预测网络模型使用具有VGG架构的卷积神经网络，通过采集患者处于昏迷状态下不同持续时长的脑电波参数，例如昏迷12小时、 24小时，训练形成不同持续时长的子评估网络模型。将患者脑电图的波形集合作为输入数据，输入到预测网络模型中进行评估，其中，针对患者处于昏迷状态下不同时长，需要使用对应的子评估网络模型作为评估规则，比如，患者已处于昏迷状态12小时，则对应使用昏迷12小时的子评估网络模型。此外还可将患者处于昏迷状态下不同持续时长的波形集合作为后续神经网络训练的脑电图数据。
其中，步骤S104基于预存的多种波形评估规则，对波形集合进行评估，包括：
S1041：基于波形评估规则包含的参考波形与波形集合中的每组特征波形组中的多个相关波形段进行匹配，获得每个相关波形段和与其对应的参考波形之间的第一匹配度；
S1042：根据第一匹配度，获得每组特征波形组的第二匹配度；
S1043：根据第二匹配度，获得波形集合的评估得分。
步骤S1041中预存的多种波形评估规则均包含每个特征波形的参考波形及其相关信息，将参考波形作为评估的参考标准，将每组特征波形组中的每个相关波形段都与其相同类别的参考波形进行匹配，并记录每个相关波形段的匹配度，作为第一匹配度。在步骤S1042中，在每组特征波形组中的每个相关波形段都匹配完毕后，综合所有第一匹配度，获得每组特征波形组的第二匹配度。步骤S1043综合每组特征波形组的第二匹配度，获得波形集合在不同评估规则下的评估得分。
步骤S105中，输出波形集合经过每种分级方法评估的预后评估结果，合并输出的结果带有预测证据，提高了预后评估的可靠性。
其中，波形集合对应的预后评估结果包括预后评价、特征波形在脑电图数据中的位置和评估得分，其中，预后评价包括预后状况好和预后状况差。
将波形集合作为输入，经过预存的多种波形评估规则评估后，合并输出每种评估规则的预后评估结果，其中预后评估结果包括评估得分，还提供相关特征波形在脑电图数据中的位置和原图，为医生查看结果提供索引，便于医生在分析患者情况时锁定脑电图中的信息，找到患病原因。预后评价有预后状况好与预后状况差两种标准，以提供给医生最简明的评估结果，便于医生判断后续对患者采用不同程度的治疗方式。比如，如果预后评价为预后状况好，则在治疗的时候可以采用更加积极的方案，以帮助患者尽快恢复。如果预后评价为预后状况差，则在治疗的时候可以尽量使用价格便宜的药品，保守治疗，以帮助患者减轻经济负担。
本公开示例性的实施例提供了一种基于脑电图的急性意识障碍预后评估装置。如图2所示，本公开中示出的基于脑电图的急性意识障碍预后评估装置的框图。该框图包括：接收模块21、提取模块22、聚类模块23、评估模块24和输出模块25。
接收模块21，被配置为获取患者的脑电图数据；
提取模块22，被配置为对脑电图数据进行识别，提取脑电图数据中的特征波形；
聚类模块23，被配置为对特征波形进行分类处理和纠错处理，形成波形集合，波形集合包括至少一组特征波形组，每组特征波形组包括多个属于同一种特征波形的多个相关波形段；
评估模块24，被配置为基于预存的多种波形评估规则，对波形集合进行评估；
输出模块25，被配置为输出每种波形评估规则下，与波形集合对应的预后评估结果。
在本公开示例性的实施例中，对特征波形进行分类处理和纠错处理，形成波形集合，聚类模块23被配置为：
将属于同一种特征波形的波形段归为一类，作为一组特征波形组，每组特征波形段中的多个波形段互为相关波形段；
对每组特征波形组中的每个相关波形段进行时频变换，获得每组特征波形组中的每个相关波形段的频域信息；
对多个频域信息进行平均值计算，将获得的平均值作为参考频域信息；
计算频域信息与参考频域信息之间的差值，将差值大于预设值的频域信息对应的相关波形段从特征波形组中剔除，形成波形集合。
在本公开示例性的实施例中，评估模块24被配置为：
基于波形评估规则包含的参考波形与波形集合中的每组特征波形组中的多个相关波形段进行匹配，获得每个相关波形段和与其对应的参考波形之间的第一匹配度；
根据第一匹配度，获得每组特征波形组的第二匹配度；
根据第二匹配度，获得波形集合的评估得分。
关于上述实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
图3是根据一示例性实施例示出的基于脑电图的急性意识障碍预后评估装置1000的框图。例如，装置1000可以是移动电话，计算机，数字广播终端，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等。
参照图3，装置1000可以包括以下一个或多个组件：处理组件1002，存储器1004，电力组件1006，多媒体组件1008，音频组件1010，输入/输出(I/O)的接口1012，传感器组件1014，以及通信组件1016。
处理组件1002通常控制装置1000的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件1002可以包括一个或多个处理器1020来执行指 令，以完成上述的方法的全部或部分步骤。此外，处理组件1002可以包括一个或多个模块，便于处理组件1002和其他组件之间的交互。例如，处理组件1002可以包括多媒体模块，以方便多媒体组件1008和处理组件1002之间的交互。
存储器1004被配置为存储各种类型的数据以支持在设备1000的操作。这些数据的示例包括用于在装置1000上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器1004可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电力组件1006为装置1000的各种组件提供电力。电力组件1006可以包括电源管理系统，一个或多个电源，及其他与为装置1000生成、管理和分配电力相关联的组件。
多媒体组件1008包括在装置1000和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件1008包括一个前置摄像头和/或后置摄像头。当设备1000处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件1010被配置为输出和/或输入音频信号。例如，音频组件1010包括一个麦克风(MIC)，当装置1000处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器1004或经由通信组件1016发送。在一些实施例中，音频组件1010还包括一个扬声器，用于输出音频信号。
I/O接口1012为处理组件1002和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件1014包括一个或多个传感器，用于为装置1000提供各个方面的状态评估。例如，传感器组件1014可以检测到设备1000的打开/关闭状态，组件的相对定位，例如组件为装置1000的显示器和小键盘，传感器组件1014还可以检测装置1000或装置1000一个组件的位置改变，用户与装置1000接触的存在或不存在，装置1000方位或加速/减速和装置 1000的温度变化。传感器组件1014可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件1014还可以包括光传感器，如CMOS或CCD图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件1014还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件1016被配置为便于装置1000和其他设备之间有线或无线方式的通信。装置1000可以接入基于通信标准的无线网络，如WiFi，2G或3G，或它们的组合。在一个示例性实施例中，通信组件1016经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，通信组件1016还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，装置1000可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述方法。
在示例性实施例中，还提供了一种包括指令的非临时性计算机可读存储介质，例如包括指令的存储器1004，上述指令可由装置1000的处理器1020执行以完成上述方法。例如，非临时性计算机可读存储介质可以是ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
一种非临时性计算机可读存储介质，当存储介质中的指令由电子设备的处理装置的处理器执行时，使得电子设备的处理装置能够执行本公开示例性的实施例所提供的是基于脑电图的急性意识障碍预后评估方法。
本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本发明的其它实施方案。本申请旨在涵盖本发明的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本发明的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本发明的真正范围和精神由下面的权利要求指出。
应当理解的是，本发明并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本发明的范围仅由所附的权利要求来限制。
工业实用性
本公开实施例通过提取脑电图中急性意识障碍的相关特征波形，形成波形集合，将现有量表和预测网络模型转换成评估规则，对波形集合中的波形进行分析，输出带有预测证据的 预后评估结果，可以极大的简化操作过程，实现自动化分级，显著提升急性意识障碍患者预后评估的可靠性、可解释性和准确性。

Claims (10)

  1. 一种基于脑电图的急性意识障碍预后评估方法，其特征在于，包括：
    获取患者的脑电图数据；
    对所述脑电图数据进行识别，提取所述脑电图数据中的特征波形；
    对所述特征波形进行分类处理和纠错处理，形成波形集合，所述波形集合包括至少一组特征波形组，每组所述特征波形组包括多个属于同一种所述特征波形的多个相关波形段；
    基于预存的多种波形评估规则，对所述波形集合进行评估；
    输出每种所述波形评估规则下，与所述波形集合对应的预后评估结果。
  2. 根据权利要求1所述的基于脑电图的急性意识障碍预后评估方法，其特征在于，所述特征波形包括α节律、θ节律、纺锤波、δ波、爆发抑制、电静息、癫痫样放电、三相波和/或GPD。
  3. 根据权利要求1所述的基于脑电图的急性意识障碍预后评估方法，其特征在于，对所述特征波形进行分类处理和纠错处理，形成波形集合，包括：
    将属于同一种所述特征波形的波形段归为一类，作为一组所述特征波形组，每组所述特征波形段中的多个波形段互为相关波形段；
    对每组所述特征波形组中的每个相关波形段进行时频变换，获得每组所述特征波形组中的每个相关波形段的频域信息；
    对多个所述频域信息进行平均值计算，将获得的平均值作为参考频域信息；
    计算所述频域信息与所述参考频域信息之间的差值，将所述差值大于预设值的所述频域信息对应的所述相关波形段从所述特征波形组中剔除，形成所述波形集合。
  4. 根据权利要求1所述的基于脑电图的急性意识障碍预后评估方法，其特征在于，所述预存的多种波形评估规则包括量表评估规则和预测网络模型评估规则。
  5. 根据权利要求4所述的基于脑电图的急性意识障碍预后评估方法，其特征在于，所述量表评估规则包括Synek、Yang、Hofmeijer中的至少一种；和/或，
    所述预测网络模型评估规则包括多个子评估网络模型，所述子评估网络模型基于患者处于昏迷状态的脑电图参数训练形成，不同的所述子评估网络模型对应的患者处于昏迷状态的持续时长不同。
  6. 根据权利要求1所述的基于脑电图的急性意识障碍预后评估方法，其特征在于，基于 预存的多种波形评估规则，对所述波形集合进行评估，包括：
    基于所述波形评估规则包含的参考波形与所述波形集合中的每组所述特征波形组中的多个相关波形段进行匹配，获得每个相关波形段和与其对应的所述参考波形之间的第一匹配度；
    根据所述第一匹配度，获得每组所述特征波形组的第二匹配度；
    根据所述第二匹配度，获得所述波形集合的评估得分。
  7. 根据权利要求6所述的基于脑电图的急性意识障碍预后评估方法，其特征在于，所述波形集合对应的预后评估结果包括预后评价、所述特征波形在所述脑电图数据中的位置和所述评估得分，其中，所述预后评价包括预后状况好和预后状况差。
  8. 一种基于脑电图的急性意识障碍预后评估装置，其特征在于，所述基于脑电图的急性意识障碍预后评估装置包括：
    接收模块，被配置为获取患者的脑电图数据；
    提取模块，被配置为对所述脑电图数据进行识别，提取所述脑电图数据中的特征波形；
    聚类模块，被配置为对所述特征波形进行分类处理和纠错处理，形成波形集合，所述波形集合包括至少一组特征波形组，每组所述特征波形组包括多个属于同一种所述特征波形的多个相关波形段；
    评估模块，被配置为基于预存的多种波形评估规则，对所述波形集合进行评估；
    输出模块，被配置为输出每种所述波形评估规则下，与所述波形集合对应的预后评估结果。
  9. 一种基于脑电图的急性意识障碍预后评估装置，其特征在于，包括：
    处理器；
    用于存储处理器可执行指令的存储器；
    其中，所述处理器被配置为执行如权利要求1-7任一项所述的基于脑电图的急性意识障碍预后评估方法。
  10. 一种非临时性计算机可读存储介质，其特征在于，当所述存储介质中的指令由基于脑电图的急性意识障碍预后评估装置的处理器执行时，使得处理器能够执行如权利要求1-7任一项所述的基于脑电图的急性意识障碍预后评估方法。
PCT/CN2022/114126 2022-08-23 2022-08-23 基于脑电图的急性意识障碍预后评估方法、装置、介质 WO2024040423A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/114126 WO2024040423A1 (zh) 2022-08-23 2022-08-23 基于脑电图的急性意识障碍预后评估方法、装置、介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/114126 WO2024040423A1 (zh) 2022-08-23 2022-08-23 基于脑电图的急性意识障碍预后评估方法、装置、介质

Publications (1)

Publication Number Publication Date
WO2024040423A1 true WO2024040423A1 (zh) 2024-02-29

Family

ID=90012156

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/114126 WO2024040423A1 (zh) 2022-08-23 2022-08-23 基于脑电图的急性意识障碍预后评估方法、装置、介质

Country Status (1)

Country Link
WO (1) WO2024040423A1 (zh)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130060158A1 (en) * 2008-05-14 2013-03-07 The Hospital For Sick Children Methods based on fluctuations in cortical synchronization
US20130245422A1 (en) * 2010-06-22 2013-09-19 National Research Council Of Canada Cognitive Function Assessment in a Patient
CN104427932A (zh) * 2012-01-18 2015-03-18 布赖恩斯科普公司 用于多模式神经评估的方法和设备
CN105147281A (zh) * 2015-08-25 2015-12-16 上海医疗器械高等专科学校 便携式意识障碍刺激促醒与评估系统
CN109893125A (zh) * 2019-03-18 2019-06-18 杭州电子科技大学 一种基于脑区信息交互的脑昏迷状态识别方法
CN111671445A (zh) * 2020-04-20 2020-09-18 广东食品药品职业学院 一种意识障碍程度分析方法
CN113116306A (zh) * 2021-04-21 2021-07-16 复旦大学 一种基于听觉诱发脑电信号分析的意识障碍辅助诊断系统
CN113598790A (zh) * 2021-07-13 2021-11-05 杭州电子科技大学 基于听觉刺激的意识障碍脑功能网络的意识评估方法
CN113712507A (zh) * 2020-05-25 2021-11-30 中国科学院脑科学与智能技术卓越创新中心 评估意识障碍程度的系统、恢复倾向预测方法和存储介质
WO2022034802A1 (ja) * 2020-08-13 2022-02-17 学校法人東北工業大学 脳から取得される波形データに対してバースト解析を行う方法、コンピュータシステム、プログラム、ならびに、バースト解析を用いて標的の状態を予測する方法、コンピュータシステム、プログラム

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130060158A1 (en) * 2008-05-14 2013-03-07 The Hospital For Sick Children Methods based on fluctuations in cortical synchronization
US20130245422A1 (en) * 2010-06-22 2013-09-19 National Research Council Of Canada Cognitive Function Assessment in a Patient
CN104427932A (zh) * 2012-01-18 2015-03-18 布赖恩斯科普公司 用于多模式神经评估的方法和设备
CN105147281A (zh) * 2015-08-25 2015-12-16 上海医疗器械高等专科学校 便携式意识障碍刺激促醒与评估系统
CN109893125A (zh) * 2019-03-18 2019-06-18 杭州电子科技大学 一种基于脑区信息交互的脑昏迷状态识别方法
CN111671445A (zh) * 2020-04-20 2020-09-18 广东食品药品职业学院 一种意识障碍程度分析方法
CN113712507A (zh) * 2020-05-25 2021-11-30 中国科学院脑科学与智能技术卓越创新中心 评估意识障碍程度的系统、恢复倾向预测方法和存储介质
WO2022034802A1 (ja) * 2020-08-13 2022-02-17 学校法人東北工業大学 脳から取得される波形データに対してバースト解析を行う方法、コンピュータシステム、プログラム、ならびに、バースト解析を用いて標的の状態を予測する方法、コンピュータシステム、プログラム
CN113116306A (zh) * 2021-04-21 2021-07-16 复旦大学 一种基于听觉诱发脑电信号分析的意识障碍辅助诊断系统
CN113598790A (zh) * 2021-07-13 2021-11-05 杭州电子科技大学 基于听觉刺激的意识障碍脑功能网络的意识评估方法

Similar Documents

Publication Publication Date Title
Hovsepian et al. cStress: towards a gold standard for continuous stress assessment in the mobile environment
US20210145306A1 (en) Managing respiratory conditions based on sounds of the respiratory system
AU2019246815B2 (en) Systems and methods for diagnosing sleep
US20130035579A1 (en) Methods for modeling neurological development and diagnosing a neurological impairment of a patient
WO2021208656A1 (zh) 睡眠风险预测方法、装置和终端设备
US20180279935A1 (en) Method and system for detecting frequency domain cardiac information by using pupillary response
CN108652587A (zh) 一种认知功能障碍预防监测装置
CN111568445B (zh) 一种基于谵妄动态预测模型的谵妄风险监测方法及系统
Ali et al. Spatio-temporal attention and magnification for classification of Parkinson’s disease from videos collected via the internet
Newman et al. Automatic nystagmus detection and quantification in long-term continuous eye-movement data
CA3189784A1 (en) Learning representations of eeg signals with self-supervised learning
CN111012345A (zh) 眼部疲劳度检测系统及方法
US11670423B2 (en) Method and system for early detection of neurodegeneration using progressive tracking of eye-markers
US10631727B2 (en) Method and system for detecting time domain cardiac parameters by using pupillary response
CN114267444A (zh) 利用睡眠结构检测阻塞性呼吸暂停及夜间额叶癫痫的方法
CN108348157A (zh) 利用多用途电容式触摸传感器的心率检测
WO2024040423A1 (zh) 基于脑电图的急性意识障碍预后评估方法、装置、介质
CN112328072A (zh) 一种基于脑电和眼电的多模态字符输入系统及方法
TWI487503B (zh) 自動睡眠分期裝置
TW202108077A (zh) 專注力自動偵測方法和系統
WO2021029526A1 (en) Methods and systems for pulmonary condition assessment
Kwon et al. Unobtrusive monitoring of ECG-derived features during daily smartphone use
CN112890830A (zh) 一种基于睡眠脑网络的抑郁症患者数据分类方法及装置
Luo et al. Exploring adaptive graph topologies and temporal graph networks for eeg-based depression detection
Guul et al. Portable prescreening system for sleep apnea

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22955972

Country of ref document: EP

Kind code of ref document: A1