Alvi, Ashik Mostafa (2023) Efficient Detection of Mild Cognitive Impairment and Alzheimer’s Disease from Brain Signal Data. PhD thesis, Victoria University.

Abstract

Brain signal data are recordings of the electrical activity of the brain made using the electroencephalography (EEG). EEG is considered the future of neuroscience, as it has emerged as the latest gold standard for detecting neurological disorders such as dementia, mild cognitive impairment (MCI), Alzheimer's disease (AD), Parkinson's disease, schizophrenia, epilepsy, and so on. Due to its cost-effectiveness and portability, EEG is becoming the first choice when it comes to analyses and the detection of neurodegenerative disorders. In addition, it has been widely accepted that analyzing EEG data is the better method for solving the challenge of learning about the brain's dynamics. EEG measures the electrical activity of the brain in real-time and can provide valuable information about brain function and dysfunction. The ability to analyze EEG data is crucial for improving our knowledge of cognitive processes and aiding in the identification of brain illnesses. To discover the most recent trends and gaps in the EEG research, a lot of scholarly articles have been studied. However, the following research problems are still existing: • Poor classification performances due to old and shallow traditional machine learning (TML) algorithms. • Lack of efficient noise removal algorithms being used. • Computationally expensive models are being proposed, which consume more power and time to diagnose brain diseases. • Limit the diagnosis to detect one disorder, even though there may be more than one brain disorder with similar symptoms. • Failed to accommodate a huge volume of EEG data. In this dissertation, all five of the aforementioned research challenges have been answered. To address the aforementioned research issues, this dissertation focuses on the detection of AD and MCI among 600 neuro-diseases. The main objective of this dissertation is to develop computer-aided diagnostic methods for effectively detecting both AD and MCI handling a large volume of EEG data with improved performance and cost-effective deep learning (DL) models.

Search Google Scholar

Repository staff login