• B. JOTHI
• Dr. D. JAYARAJ

we have a tendency to propose a replacement approach for the detection of Epileptical victimization by image processing and deep learning methodology. A profound study of the EEG (EEG) recordings is needed for the correct detection of these epileptic seizures. Therefore, we propose a model that provides reliable strategies of each preprocessing and extraction. The foremost step concerned during this paper is obtaining EEG signals as input and generating corresponding brain map and applying ICA as pre-processor to induce filtered image. The LBP is used to extract pattern options from filtered image and the proposed PATCH MEAN ASSISTED deep learning (PMA-DL) are used for classifying Epileptical seizure. Here the input image is split into patches. Every patch is split into 2 layers “Convolutional layer ” and “Pooling layer” then repetition the steps for whole image for generating test features and to extract applied mathematics options of the raw signal, like finding mean values and obtaining weighted mean for the image. All the features are combined and therefore the dimensions of the embedded feature vector are reduced by the principal component analysis (PCA). Finally, the least square support vector machine (LSSVM) classifier is employed to classify the seizure. The planned system is achieved accuracy 97%, Sensitivity 98.1%, and Specificity 99.9%, in classifying the patients which is 1% higher than the existing methods.

Convolutional Neural Network(CNN), epileptic seizure, EEG brain map, Least Square Support Vector Machine(LSSVM), Local Binary Pattern (LBP), Patch Mean Assisted-Deep Learning(PMA- DL), Principal Component Analysis(PCA).

"ADVANCE ARTIFICAL INTELLIGENCE PRACTICE FOR PAROXYSM SEIZURE DETECTION ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 2, page no.f171-f176, February-2024, Available :http://www.jetir.org/papers/JETIR2402520.pdf

"ADVANCE ARTIFICAL INTELLIGENCE PRACTICE FOR PAROXYSM SEIZURE DETECTION ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 2, page no. ppf171-f176, February-2024, Available at : http://www.jetir.org/papers/JETIR2402520.pdf