Machine Learning-Based Brain Diseases Diagnosing in Electroencephalogram Signals, Alzheimer’S, and Parkinson’S Publisher



Tavakoli N¹ ; Karimi Z² ; Asadijouzani S³ ; Azizi N⁴ ; Rezakhani S⁵ ; Tobeiha A⁵ Show Affiliations
Source: Studies in Big Data Published:2022

Abstract

Brain disfunction is very common in old age and even in middle-aged people. Alzheimer’s and Parkinson’s diseases are among the most common diseases due to brain disfunction. In Alzheimer’s disease, a person gradually loses his mental abilities. Although it is normal for people to become a little forgetful as they get older, this memory disorder gradually progresses, posing great challenges. In order to prevent the spread of Alzheimer’s disease, early detection will be very helpful. Parkinson’s is another disease that will increase in prevalence as life expectancy increases. Brain monitoring tools are used to detect these diseases early. An inexpensive and useful tool and low-risk brain signals are electroencephalograms. In order to analyze brain signals, the use of machine learning-based methods has been able to show its superiority. In order to diagnose Alzheimer’s and Parkinson's in machine learning, there are preprocessing steps, feature extraction, feature selection, classification, and evaluation. Since electroencephalogram data have high repetition and correlation in different channels recorded on the head, feature extraction techniques will be of great importance. Feature selection methods seek to select the most effective features to classify and identify disease status. Finally, the selected features will be categorized using different categories. In this chapter, a complete overview of the stages of diagnosis of these diseases with the help of machine learning will be provided. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.