Browsing by Author "MUHAMMAD HAMZA CHAUDARY"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Detection of epilepsy through deep learning
    (UMT.Lahore, 2024) MUHAMMAD MUSTAFA AWAN; MUHAMMAD HAMZA CHAUDARY
    Detecting and classifying epileptic seizures is crucial in the analysis, monitoring, and diagnosis of patients with epilepsy, especially when it comes to realizing and actualizing computer-aided devices. EEG system and pre-processing the data to remove artifacts and enhance the ratio of signal to noise. From the pre-processed EEG data, different fea- tures are derived, including but not limited to time-domain features, frequency-domain features, and statistical features. The accuracy of the detection and classification pro- cess plays a vital role in determining the effectiveness of these applications. Over the years, numerous approaches have been explored, suggested, and advanced to enhance the precision of these methods. In this report, the goal is to develop an Artificial- Intelligence-based epilepsy diagnosis tool that accurately classifies all subtypes under the Generalized Epilepsy (GE) and Focal Epilepsy (FE) with minimal computational complexity. To achieve this goal, the study investigates various seizure detection and classification algorithms, including the latest deep learning algorithms. The proposed model consists of a feature extraction layer that extracts critical features from the incom- ing EEG signals and a classification layer that uses the state-of-the-art Deep Learning algorithms to classify the incoming signals into its types using weighted features. The experimental findings demonstrate that the suggested model can accurately classify all eight types of epileptic seizures. The proposed model achieves an average precision of 0.933, an average specificity of 0.994, and an average F1 score of 0.876, demonstrating its high accuracy in the classification of seizures