Chapter 1 - Early detection of neurological diseases using machine learning and deep learning techniques: A review

Abstract

Alzheimer's disease (AD) and Parkinson's disease (PD) are two common neurological diseases that affect millions of people older than 65years. PD is a progressive neurological disorder characterized by motor and nonmotor syndromes that affect the daily lives of patients. Individuals with PD experience difficulty speaking, writing, walking, or completing other simple tasks due to damage or death of the part of the brain responsible for producing dopamine. As it is a chronic disease, PD symptoms grow worse over time. Generally, medical observations and assessment of clinical signs are used to diagnose PD and AD. However, traditional diagnostic approaches may suffer from subjectivity, as they rely on the evaluation of movements that are sometimes subtle and therefore difficult to classify, leading to possible misdiagnosis. In the meantime, early nonmotor symptoms of PD may be mild and can be caused by many other conditions. Therefore, these symptoms are often ignored, making diagnosis of PD at an early stage challenging. AD is a common form of dementia that doesn’t affect patients during the early stages. Early diagnosis of AD is very difficult because patients are usually asymptomatic in the early stages. Thus, many researchers are studying AD in an attempt to develop successful methods of early diagnosis. To address these difficulties and refine the diagnosis and assessment procedures of PD and AD, machine learning (ML) and deep learning (DL) methods have been implemented. ML and DL are subfields of artificial intelligence (AI) that are increasingly applied to several medical diagnosis tasks. Despite the good performance of hand-crafted ML algorithms, there is still a problem linked to feature extraction and selection. DL, however, has provided a solution to this issue. Several studies on the diagnosis of PD and AD using DL algorithms have recently been conducted. In this chapter, we provide an overview of the application of hand-crafted ML algorithms and DL techniques for diagnosis of both diseases. In addition, we propose an edge-based healthcare system architecture for monitoring AD patients by using the accelerometer sensor of a smartphone.