scholarly journals Localizing Epileptogenic Zone from High Density EEG Data Using Machine Learning

2021 ◽  
Vol 17 (06) ◽  
pp. 73
Author(s):  
Sehresh Khan ◽  
Aunsia Khan ◽  
Nazia Hameed ◽  
Muhammad Aleem Taufiq ◽  
Saba Riaz
Keyword(s):  
Machine Learning ◽  
Human Brain ◽  
Antiepileptic Drugs ◽  
Focal Epilepsy ◽  
Brain Regions ◽  
High Density ◽  
Machine Learning Algorithms ◽  
Epileptogenic Zone ◽  
Drug Resistant ◽  
Time Frequency

<span>Drug-resistant focal epilepsy is the failure of antiepileptic drugs scheduled to obtain epileptic free brain activities. In human brain, cerebral hemispheres are the most commonly involved brain regions in epilepsy. In case of antiepileptic drugs failure, surgical treatment is the best cure possible. However, correct localization of epileptogenic region is a challenging task for neurologists, while for computer scientists, automatic localization is. This research work’s aim is to explore the functional activities of all brain regions in drug-resistant focal epileptic patients and achieve high accuracy for the classification of epileptogenic region (ER) with the high-density electroencephalographic (hdEEG) data. The proposed system includes frequency analysis for feature extractions followed by individual subject’s registration of hdEEG signals with anatomical brain images for most precise localization of ER possible. The datasets attained from feature extraction process are then preprocessed for class imbalanced and then evaluated using different machine learning algorithms including the techniques under Bayesian networks, Lazy networks, Meta techniques, Rule based systems and Tree structured algorithms. Considering human brain as stationary object as well as dynamic object, frequency-based and time-frequency based features were considered in 12 subjects respectively. Through this novel approach, 99.70% accuracy is achieved to classify ER from healthy regions using KSTAR and using IBK algorithm, 91.60% accuracy has been achieved to classify generator from propagator.</span>

scholarly journals Comparison of Resampling Techniques for Imbalanced Datasets in Machine Learning: Application to Epileptogenic Zone Localization From Interictal Intracranial EEG Recordings in Patients With Focal Epilepsy

2021 ◽  
Vol 15 ◽  
Author(s):  
Giulia Varotto ◽  
Gianluca Susi ◽  
Laura Tassi ◽  
Francesca Gozzo ◽  
Silvana Franceschetti ◽  
...  
Keyword(s):  
Machine Learning ◽  
Focal Epilepsy ◽  
Geometric Mean ◽  
Ensemble Methods ◽  
Classification Problem ◽  
Brain Regions ◽  
Classification Method ◽  
Machine Learning Techniques ◽  
Epileptogenic Zone ◽  
Original Dataset

Aim: In neuroscience research, data are quite often characterized by an imbalanced distribution between the majority and minority classes, an issue that can limit or even worsen the prediction performance of machine learning methods. Different resampling procedures have been developed to face this problem and a lot of work has been done in comparing their effectiveness in different scenarios. Notably, the robustness of such techniques has been tested among a wide variety of different datasets, without considering the performance of each specific dataset. In this study, we compare the performances of different resampling procedures for the imbalanced domain in stereo-electroencephalography (SEEG) recordings of the patients with focal epilepsies who underwent surgery.Methods: We considered data obtained by network analysis of interictal SEEG recorded from 10 patients with drug-resistant focal epilepsies, for a supervised classification problem aimed at distinguishing between the epileptogenic and non-epileptogenic brain regions in interictal conditions. We investigated the effectiveness of five oversampling and five undersampling procedures, using 10 different machine learning classifiers. Moreover, six specific ensemble methods for the imbalanced domain were also tested. To compare the performances, Area under the ROC curve (AUC), F-measure, Geometric Mean, and Balanced Accuracy were considered.Results: Both the resampling procedures showed improved performances with respect to the original dataset. The oversampling procedure was found to be more sensitive to the type of classification method employed, with Adaptive Synthetic Sampling (ADASYN) exhibiting the best performances. All the undersampling approaches were more robust than the oversampling among the different classifiers, with Random Undersampling (RUS) exhibiting the best performance despite being the simplest and most basic classification method.Conclusions: The application of machine learning techniques that take into consideration the balance of features by resampling is beneficial and leads to more accurate localization of the epileptogenic zone from interictal periods. In addition, our results highlight the importance of the type of classification method that must be used together with the resampling to maximize the benefit to the outcome.

scholarly journals Hierarchical Tactile Sensation Integration from Prosthetic Fingertips Enables Multi-Texture Surface Recognition

Sensors ◽  
2021 ◽  
Vol 21 (13) ◽  
pp. 4324
Author(s):  
Moaed A. Abd ◽  
Rudy Paul ◽  
Aparna Aravelli ◽  
Ou Bai ◽  
Leonel Lagos ◽  
...  
Keyword(s):  
Machine Learning ◽  
Learning Algorithms ◽  
Machine Learning Algorithms ◽  
Sliding Contact ◽  
Tactile Sensation ◽  
Support Vector ◽  
Textured Surfaces ◽  
K Nearest Neighbor ◽  
Tactile Sensors ◽  
Time Frequency

Multifunctional flexible tactile sensors could be useful to improve the control of prosthetic hands. To that end, highly stretchable liquid metal tactile sensors (LMS) were designed, manufactured via photolithography, and incorporated into the fingertips of a prosthetic hand. Three novel contributions were made with the LMS. First, individual fingertips were used to distinguish between different speeds of sliding contact with different surfaces. Second, differences in surface textures were reliably detected during sliding contact. Third, the capacity for hierarchical tactile sensor integration was demonstrated by using four LMS signals simultaneously to distinguish between ten complex multi-textured surfaces. Four different machine learning algorithms were compared for their successful classification capabilities: K-nearest neighbor (KNN), support vector machine (SVM), random forest (RF), and neural network (NN). The time-frequency features of the LMSs were extracted to train and test the machine learning algorithms. The NN generally performed the best at the speed and texture detection with a single finger and had a 99.2 ± 0.8% accuracy to distinguish between ten different multi-textured surfaces using four LMSs from four fingers simultaneously. The capability for hierarchical multi-finger tactile sensation integration could be useful to provide a higher level of intelligence for artificial hands.

scholarly journals Stereoelectroencephalography-Guided Radiofrequency Thermocoagulation (SEEG-Guided RF-TC) in Patients with Drug-Resistant Focal Epilepsy

2017 ◽  
Vol 3 (1) ◽  
pp. 40-47
Author(s):  
Chengwei Xu ◽  
Wenjing Zhou
Keyword(s):  
Radio Frequency ◽  
Surgical Resection ◽  
Seizure Frequency ◽  
Palliative Treatment ◽  
Focal Epilepsy ◽  
Epileptogenic Zone ◽  
Drug Resistant ◽  
Functional Area ◽  
Presurgical Evaluation ◽  
Epileptic Patients

For some patients with drug-resistant focal epilepsy, we usually select conventional surgical resection, which has brought better outcomes. However, others are not eligible for a conventional open surgical resection of the epileptogenic zone because of the proximity of a functional area or the implication of a larger epileptogenic network. Initially, stereoelectroencephalography (SEEG) exploration was a method of electroencephalography recording that was used in the presurgical evaluation of epileptic patients with complex epilepsy. Later, intracerebral electrodes used for SEEG were applied to produce radio frequency thermocoagulation (RF-TC) in epileptic patients. SEEG-guided RF-TC has produced some promising results, especially in the last dacade. Now, it has become popular as a palliative treatment to reduce seizure frequency in patients with drug-resistant focal epilepsy. This article presents a review of SEEG-guided RF-TC.

scholarly journals Role of Repetitive Transcranial Magnetic Stimulation on Drug Resistant Focal Epilepsy: A Controlled Clinical Study

QJM ◽  
2020 ◽  
Vol 113 (Supplement_1) ◽  
Author(s):  
S Ashour ◽  
A Gaber ◽  
T Hussein ◽  
A Kamal
Keyword(s):  
Transcranial Magnetic Stimulation ◽  
Magnetic Stimulation ◽  
Focal Epilepsy ◽  
Repetitive Transcranial Magnetic Stimulation ◽  
Low Frequency ◽  
Baseline Period ◽  
Epileptogenic Zone ◽  
Drug Resistant ◽  
Resting Motor Threshold ◽  
Before And After

Abstract Copyright 2019 Department of Neuropsychiatry, Ain Shams University. Unauthorized reproduction of this article is prohibited Purpose This study was designed to evaluate the therapeutic effect of low-frequency repetitive transcranial magnetic stimulation (rTMS) on patients with drug resistant focal epilepsy. Methods Fifty-five patients with drug resistant neocortical focal epilepsy were screened and 30 patients were divided into two groups (active and sham ) were given (0.5 hz , 1000 pulses , 90%of resting motor threshold (rMT)) on stimulation site detected by correlation between clinical semiology, EEG and or MRI finding. Seizures frequency, severity and seizure free days were compared for one month before and after rTMS with no change in antiepileptic drugs. We assumed 50% seizure reductionrate after rTMS. Results Seizures significantly decreased following rTMS treatment 50% seizure reduction was achieved 87.5%inthe active rTMS groupand50% in the sham rTMS group(p &gt; 0.03).in the follow-up period compared to baseline period. Seizure free days %increase in active group 20.7% compared to sham group 7.5% (p = 0.0501). Significance Low-frequency rTMS delivered into the epileptogenic zone had a significant antiepileptic effect on patients with drug resistant neocortical focal epilepsy. © 2018 Department of Neuropsychiatry, Ain Shams University

scholarly journals fMRI-based Decoding of Visual Information from Human Brain Activity: A Brief Review

2021 ◽  
Author(s):  
Shuo Huang ◽  
Wei Shao ◽  
Mei-Ling Wang ◽  
Dao-Qiang Zhang
Keyword(s):  
Machine Learning ◽  
Human Brain ◽  
Visual Information ◽  
Pattern Analysis ◽  
Brain Activity ◽  
Machine Learning Algorithms ◽  
Fmri Data ◽  
Future Research ◽  
Neural Activation ◽  
Research Problems

AbstractOne of the most significant challenges in the neuroscience community is to understand how the human brain works. Recent progress in neuroimaging techniques have validated that it is possible to decode a person’s thoughts, memories, and emotions via functional magnetic resonance imaging (i.e., fMRI) since it can measure the neural activation of human brains with satisfied spatiotemporal resolutions. However, the unprecedented scale and complexity of the fMRI data have presented critical computational bottlenecks requiring new scientific analytic tools. Given the increasingly important role of machine learning in neuroscience, a great many machine learning algorithms are presented to analyze brain activities from the fMRI data. In this paper, we mainly provide a comprehensive and up-to-date review of machine learning methods for analyzing neural activities with the following three aspects, i.e., brain image functional alignment, brain activity pattern analysis, and visual stimuli reconstruction. In addition, online resources and open research problems on brain pattern analysis are also provided for the convenience of future research.

scholarly journals In-process detection of grinding burn using machine learning

2021 ◽  
Author(s):  
Emil Sauter ◽  
Erkut Sarikaya ◽  
Marius Winter ◽  
Konrad Wegener
Keyword(s):  
Machine Learning ◽  
Machine Learning Algorithms ◽  
Limiting Factors ◽  
Process Time ◽  
New Approach ◽  
Time Frequency ◽  
Grinding Burn ◽  
Process Measurement ◽  
Feature Selector ◽  
Process Detection

AbstractThe improvement of industrial grinding processes is driven by the objective to reduce process time and costs while maintaining required workpiece quality characteristics. One of several limiting factors is grinding burn. Usually applied techniques for workpiece burn are conducted often only for selected parts and can be time consuming. This study presents a new approach for grinding burn detection realized for each ground part under near-production conditions. Based on the in-process measurement of acoustic emission, spindle electric current, and power signals, time-frequency transforms are conducted to derive almost 900 statistical features as an input for machine learning algorithms. Using genetic programming, an optimized combination between feature selector and classifier is determined to detect grinding burn. The application of the approach results in a high classification accuracy of about 99% for the binary problem and more than 98% for the multi-classdetection case, respectively.

scholarly journals What Do Machines Tell us About Dementia? Machine Learning Applied to Aging, Dementia and Traumatic Brain Injury Study

2021 ◽  
Author(s):  
Denis Arthur Pinheiro Moura ◽  
Joao Ricardo Mendes de Oliveira
Keyword(s):  
Gene Expression ◽  
Machine Learning ◽  
Traumatic Brain Injury ◽  
Brain Injury ◽  
Expression Profiles ◽  
Brain Regions ◽  
Machine Learning Algorithms ◽  
Differentially Expressed ◽  
Classification Model ◽  
Specific Gene

Abstract Dementia, a syndrome characterized by the progressive deterioration of memory and cognition, arises from different pathologies, with Alzheimer's Disease (AD) its most common cause. Patterns of gene expression during dementia of different etiologies may function as generalist biomarkers of the condition. We used RNA-Seq data from the Allen Dementia and Traumatic Brain Injury Study (ADTBI) to identify differentially expressed genes in brains with dementia. Machine Learning algorithms Decision Trees (DT) and Random Forest (RF) were used to create models to identify dementia samples based on their gene expression profile. Importance analyses were conducted to identify the most relevant genes in each classification model. A total of 1629 differentially expressed (DE) genes were found in brains with the condition. Gene PAN3-AS1 was the only DE gene across more than three brain regions. The artificial intelligence models were capable of identifying correctly up to 92.85% of dementia samples. Our analyses provide interesting insights regarding using brain-specific gene expression profiles as biomarkers of dementia, identifying genes possibly involved with dementia, and guiding future studies in prediction and early identification of the syndrome.

scholarly journals Efficacy and tolerability of rufinamide in the treatment of epilepsy (experience of the Svt. Luka’s Institute of Child Neurology and Epilepsy)

2018 ◽  
Vol 13 (2) ◽  
pp. 7-19
Author(s):  
K. Yu. Mukhin ◽  
O. A. Pylaeva ◽  
M. Yu. Bobylova ◽  
N. V. Freydkova ◽  
L. Yu. Glukhova ◽  
...  
Keyword(s):  
Drug Resistance ◽  
Side Effects ◽  
Antiepileptic Drugs ◽  
Seizure Frequency ◽  
Focal Epilepsy ◽  
Alternative Methods ◽  
Chromosomal Microarray ◽  
Study Cohort ◽  
Chromosomal Microarray Analysis ◽  
Drug Resistant

Background. Despite significant advances in epileptology, approximately one-third of epilepsy patients suffer from drug-resistant seizures. Numerous approaches are currently available to treat epilepsy; however, there are still many patients with treatment-resistant epilepsy, in whom surgical treatment is impossible and alternative methods (vagus nerve stimulation and ketogenic diet) are ineffective. Therefore, searching for novel effective antiepileptic drugs (AEDs) is crucial for these patients.Objective: analysis of own data on the efficacy and tolerability of rufinamide in patients with severe forms of epilepsy and seizures typical of Lennox–Gastaut syndrome (LGS).Materials and methods. The study included 31 patients aged between 4 and 26 years (mean age 7.5 years) that received rufinamide (inovelon). The study cohort comprised 21 males and 10 females. Fifteen patients were diagnosed with LGS, whereas 16 patients were diagnosed with structural focal epilepsy with a phenocopy of LGS. Five patients had an evolution of West syndrome to LGS. The majority of patients (n = 22) experienced predominantly axial tonic seizures and epileptic spasms that were considered as indications for introduction of rufinamide. All patients underwent electroencephalography, video-electroencephalography monitoring during wakefulness and sleep, magnetic resonance imaging (MRI) (including high-resolution MRI with special epilepsy protocols when indicated), genetic examination (tandem mass spectrometry, hereditary epilepsy gene panel test and chromosomal microarray analysis) when indicated, and laboratory tests to assess tolerability of antiepileptic drugs.Results. Good therapeutic effect (more than 50 % reduction in seizure frequency) was achieved in 14 (45.2 %) patients. A less than 50 % reduction in seizure frequency occurred in 5 (16.1 %) patients; in 2 of them seizures became shorter and milder without a significant reduction in their frequency. Rufinamide was ineffective in 9 (29 %) patients. Three (9.7 %) patients experienced aggravation (increased seizure frequency) after the introduction of rufinamide. Thus, treatment with rufinamide was effective in 19 (61.3 %) patients. Rufinamide was well tolerated by most of the patients. Side effects were observed in 6 (19 %) participants. Side effects (forced normalization) caused withdrawal of rufinamide in 1 (3.2 %) patient. Currently, 10 (32 %) patients continue to take rufinamide. Sixteen patients received rufinamide for <6 months, 17 patients – for >6 months, 5 patients – for >12 months, and 1 patient – for >2 years.Conclusion. Our findings are consistent with the results obtained by foreign authors in routine clinical practice. In our study, rufinamide was used only in patients with drug-resistant epilepsy that earlier received many of currently available AEDs (both in monotherapy and in combination with other drugs). All study participants were earlier treated with at least three different AEDs that were ineffective. Seven patients received more than 8 AEDs in various combinations. This initial drug resistance should be taken into account when analyzing the data, which can not be extrapolated to patients with unknown drug resistance. We assume that the early introduction of rufinamide (prior to the detection of drug resistance) might have yielded better results.

scholarly journals Local propagation dynamics of MEG interictal spikes: source reconstruction with traveling wave priors

2020 ◽  
Author(s):  
Aleksandra Kuznetsova ◽  
Mikhail Lebedev ◽  
Alexei Ossadtchi
Keyword(s):  
Traveling Wave ◽  
Large Scale ◽  
Brain Regions ◽  
Propagation Model ◽  
Sensor Data ◽  
Epileptogenic Zone ◽  
Drug Resistant ◽  
Interictal Spikes ◽  
Propagation Dynamics ◽  
Clinical Records

AbstractEpilepsy is one of the most common neurological disorders, with about 30% of cases being drug-resistant and requiring surgical intervention. To localize the epileptogenic zone (EZ), the pathological area that has to be surgically removed, brain regions are inspected for the presence of spikes during the interictal periods. This procedure maps irritative zones where spikes are present, but it is still challenging to determine which of the irritative zones generate seizures. To localize the source of seizures more precisely, a large-scale approach could be applied where the causal relationship is assessed between the signals recorded in a finite number of irritative zones [27]. This method however, does not reveal the fine-grained spatiotemporal patterns of spikes, which could provide valuable information regarding EZ location and increase the likelihood of surgery success [33].Here we present a framework to noninvasively investigate the fine patterns of interictal spikes present in magnetoencephalographic (MEG) data. We use a traveling wave model, previously employed in the analysis of cortical alpha oscillations [16], to regularize the MEG inverse problem and to determine the cortical paths of spike traveling waves. Our algorithm represents spike propagation patterns as a superposition of local waves traveling along radial paths stemming from a single origin. With the help of the positively constrained LASSO technique we scan over wave onset moment and propagation velocity parameters to determine their combination that yields the best fit to the MEG sensor data of each spike.We first used realistically simulated MEG data to validate the algorithm ability to successfully track interictal activity on a millimeter-millisecond scale. Next, we examined MEG data from three patients with drug-resistant epilepsy. Wave-like spike patterns with clear propagation dynamics were found in a fraction of spikes, whereas the other fraction could not be explained by the wave propagation model with a small number of propagation directions. Moreover, in agreement with the previous work [33], the spike waves with clear propagation dynamics exhibited spatial segregation and matched the clinical records on seizure onset zones (SOZs) available for two patients out of three.

scholarly journals ECG-based machine-learning algorithms for heartbeat classification

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Saira Aziz ◽  
Sajid Ahmed ◽  
Mohamed-Slim Alouini
Keyword(s):  
Machine Learning ◽  
Fractional Fourier Transform ◽  
Heart Diseases ◽  
Learning Model ◽  
Machine Learning Algorithms ◽  
Heartbeat Classification ◽  
Time Frequency ◽  
Ecg Signals ◽  
Signal Features ◽  
Machine Learning Model

AbstractElectrocardiogram (ECG) signals represent the electrical activity of the human hearts and consist of several waveforms (P, QRS, and T). The duration and shape of each waveform and the distances between different peaks are used to diagnose heart diseases. In this work, to better analyze ECG signals, a new algorithm that exploits two-event related moving-averages (TERMA) and fractional-Fourier-transform (FrFT) algorithms is proposed. The TERMA algorithm specifies certain areas of interest to locate desired peak, while the FrFT rotates ECG signals in the time-frequency plane to manifest the locations of various peaks. The proposed algorithm’s performance outperforms state-of-the-art algorithms. Moreover, to automatically classify heart disease, estimated peaks, durations between different peaks, and other ECG signal features were used to train a machine-learning model. Most of the available studies uses the MIT-BIH database (only 48 patients). However, in this work, the recently reported Shaoxing People’s Hospital (SPH) database, which consists of more than 10,000 patients, was used to train the proposed machine-learning model, which is more realistic for classification. The cross-database training and testing with promising results is the uniqueness of our proposed machine-learning model.

Sign in / Sign up

Export Citation Format

Share Document