Fast Detection and Quantification of Interictal Spikes and Seizures in a Rodent Model of Epilepsy Using an Automated Algorithm
使用自动化算法快速检测和量化啮齿动物癫痫模型中的发作间期尖峰和癫痫发作
(*contributed equally to this work) 发布: 2023年03月20日第13卷第6期 DOI: 10.21769/BioProtoc.4632 浏览次数: 1067
评审: Prashanth N SuravajhalaJayaraman ValadiWenyang LiAnonymous reviewer(s)
下载 PDF 
Q&A
Abstract
The electroencephalogram (EEG) is a powerful tool for analyzing neural activity in various neurological disorders, both in animals and in humans. This technology has enabled researchers to record the brain’s abrupt changes in electrical activity with high resolution, thus facilitating efforts to understand the brain’s response to internal and external stimuli. The EEG signal acquired from implanted electrodes can be used to precisely study the spiking patterns that occur during abnormal neural discharges. These patterns can be analyzed in conjunction with behavioral observations and serve as an important means for accurate assessment and quantification of behavioral and electrographic seizures. Numerous algorithms have been developed for the automated quantification of EEG data; however, many of these algorithms were developed with outdated programming languages and require robust computational hardware to run effectively. Additionally, some of these programs require substantial computation time, reducing the relative benefits of automation. Thus, we sought to develop an automated EEG algorithm that was programmed using a familiar programming language (MATLAB), and that could run efficiently without extensive computational demands. This algorithm was developed to quantify interictal spikes and seizures in mice that were subjected to traumatic brain injury. Although the algorithm was designed to be fully automated, it can be operated manually, and all the parameters for EEG activity detection can be easily modified for broad data analysis. Additionally, the algorithm is capable of processing months of lengthy EEG datasets in the order of minutes to hours, reducing both analysis time and errors introduced through manual-based processing.
Automated (自动化)
Algorithm (算法)
Interictal Spikes (发作间期尖峰)
Seizures (癫痫发作)
MATLAB (MATLAB)
Background
The emphasis of translational research in epilepsy is now shifting from the development of anti-seizure therapies to anti-epileptogenic and disease-modifying treatments. Long-term electroencephalographic recordings during the acute and chronic phases of disease are an important part of the drug-testing paradigm for these therapies. Manual analysis of prolonged electroencephalography (EEG) recordings can be time-consuming, laborious, and expensive. Further, manual EEG interpretation risks the introduction of human error and implicit bias, which can negatively impact data quantification and subsequent findings (Arends et al., 2017; Benbadis et al., 2017; Barger et al., 2019). Improper EEG analysis generates inconsistent data, which skews interpretation of the efficacy of interventional therapies at preventing the advent or progression of disease. Additionally, the limited signal-to-noise ratio makes manual analysis of seizures, seizure patterns, epileptiform spikes, and detection of other epileptiform events extremely challenging (Kaplan et al., 2005; Liu et al., 2019). These issues present a sizeable roadblock to rapid and unbiased scoring of electrographic spikes and seizure analysis. An automated, algorithm-based EEG analysis program is an ideal solution to meet this goal.
Previous work towards the development of automated EEG analysis tools has led to the production of algorithms that are capable of detecting and quantifying interictal activity and seizures, both in real-time and post data collection (Bergstrom et al., 2013;Tieng et al., 2016 and 2017). Unfortunately, many of these algorithms have been produced in programming languages that are no longer widely used or require substantial computational power to operate effectively (Harner et al., 2009). As a result of the steady evolution in the capabilities of computer processing equipment, higher levels of computational power can now be achieved. However, these new computational technologies are expensive, and often require additional hardware upgrades to be utilized in computing systems. Therefore, we developed an algorithm that detects and quantifies interictal spikes and seizures, using a programming language that requires minimal computational power and a computer with average processing hardware.
We evaluated our novel spike detection algorithm—a Basic Function Algorithm programmed in MATLAB—using large-scale EEG data obtained from a mouse model of epilepsy. We utilized EEG traces from mice with traumatic brain injury–induced epilepsy. The interictal spikes and seizures were manually detected using commercially available software (Neuroscore, Data Science International), and automatically detected using our EEG algorithm. All seizures detected using the automated algorithm were cross-referenced with Neuroscore for verification. Our comparative analysis revealed that the automated, algorithmic scoring of electrographic spikes and seizures is faster, more accurate, and less laborious than manual EEG analysis. Moreover, the algorithm is more dependable, resulting in a lower risk of introducing bias than manual, visual-based analysis.
Materials and Reagents
Materials for electrode implantation
Forceps, scalpel handles, scissors (Stoelting, catalog numbers: 5210883P, 52171P, 5210002P)
26 Gauge needles (Fisher Scientific, catalog number: BD 305111)
1 mL syringes (Fisher Scientific, catalog number: BD 309659)
Cotton tip applicators (Puritan, catalog number: 806 WC)
Burr drill tips (Stoelting, catalog number: 514552)
Surgical clips (Cellpoint Scientific, catalog number: 203-1000)
Surgical sutures, size 5-0 (Ethicon, catalog number: J493G)
Dental cement (Co-Oral-ite MFG Co, catalog number: 525000)
HD-X02 implants with two biopotential channels [Data Sciences International (DSI) (division of Harvard Bioscience, Inc)], catalog number: 270-0172-001)
Reagents
70% ethanol
Chlorhexidine scrub (Mölnlycke, catalog number: 0234-0575-08)
0.9% saline solution
Artificial tears ointment (Aventix, catalog number: 13585)
Meloxicam (Norbrook, catalog number: 0342.90A)
Baytril (Bayer Pharmaceuticals LLC, catalog number: AH039GH)
5% Haemo-Sol regular (Haemo-Sol, catalog number: 026-050)
Equipment
EEG acquisition system
Ponemah Software System (DSI, catalog number: PNM-P3P-CFG)
Included in system:
Analysis Module – Electroencephalogram Analysis Software Module (PNM-EEG100W)
Video Module – Noldus media recorder (PNM-VIDEO-008)
Ponemah Software 6.51 (PNM-P3P-651)
New Ponemah Acquisitions Software (PNM-P3P-TELELV8)
Analysis Module – Electromyogram Analysis Software module (PNM-EMG100W)
Computer – Lenovo ThinkPad T490 64-bit Windows 10 with docking station (271-0112-031)
Small Business Router – Cisco RV160 VPN router (DSI, catalog number: RV160)
Network Switch – 183w Network Switch (DSI, catalog number: 271-0115-003)
MX2 – Matrix 2.0 with USB Port for Signal Interface Support (DSI, catalog number: 271-0119-002)
RPC-1 – Receiver Pad for Plastic Cages with 4.5 m cable (DSI, catalog number: 272-6001-001)
Neuroscore software (DSI, 271-0165-CFG)
Included in system:
NS 3.3 Core – Neuroscore v3.3 Core software (271-0165-330)
Seizure Module – Neuroscore Seizure module (271-0167-SEIZURE)
Video Module – Neuroscore Video Synchronization module (271-0168-VIDEO)
Video Camera – Axis M1145-L Network Camera Kit with Axis T8120 Midspan (AXIS, catalog number: 275-0204-001)
Computational hardware
One of the primary goals of this project was to develop an autonomous algorithm that could rapidly process large volumes of EEG data using minimal computational power. To do so, an HP Spectre x360 laptop (Manufacturer: HP, Model: 15-eb1043dx) was used to process all EEG data. The CPU utilized was an Intel i7-10510U running at 2.30 GHz base clock speed, with a maximum clock speed of 4.30 GHz on all four cores. A total of 16 GB of GDDR4 memory was operated at 2667 MHz. The laptop had an additional 32 GB of Intel Optane memory available; however, this was disabled during data processing, and thus the maximum memory available at any given time was 16 GB. The laptop contained an Nvidia MX250 graphics card; however, since MATLAB is a CPU-intensive program, GPU processing power was not strongly considered. EEG data was written on an internal M.2 solid-state drive, after increased processing times were observed when using a standard USB 2.0 hard disk drive.
Software
Windows 10 Operating System (Microsoft, https://www.microsoft.com/en-us/software-download/windows10)
Neuroscore, Version 3.3 (Data Sciences International, https://www.datasci.com/products/software/neuroscore)
Ponemah Software 6.51 (https://www.datasci.com/products/software/ponemah)
MATLAB R2021a (MathWorks, https://www.mathworks.com/products/matlab.html)
MATLAB Signal Processing Toolbox (MathWorks, https://www.mathworks.com/products/signal.html)
Procedure
文章信息
版权信息
© 2023 The Author(s); This is an open access article under the CC BY-NC license (https://creativecommons.org/licenses/by-nc/4.0/).
如何引用
Jackson, K. J., Sharma, S., Tiarks, G., Rodriguez, S. and Bassuk, A. G. (2023). Fast Detection and Quantification of Interictal Spikes and Seizures in a Rodent Model of Epilepsy Using an Automated Algorithm. Bio-protocol 13(6): e4632. DOI: 10.21769/BioProtoc.4632.
Download Citation in RIS Format
分类
神经科学 > 神经系统疾病 > 癫痫
神经科学 > 基础技术
您对这篇实验方法有问题吗？
在此处发布您的问题，我们将邀请本文作者来回答。同时，我们会将您的问题发布到Bio-protocol Exchange，以便寻求社区成员的帮助。
提问指南
+ 问题描述
写下详细的问题描述，包括所有有助于他人回答您问题的信息（例如实验过程、条件和相关图像等）。
发布问题
