Brain Mapping for Dementia Disorder


Textbook, 2020

54 Pages, Grade: 2.0


Excerpt


Contents

List of Figures

List of Tables

1 Introduction

2 Neural Disorders, Dataset Description and Preliminary Signal Processing
2.1 Alzheimer’s Disease
2.1.1 Attention deficit/hyperactive disorder (ADHD)
2.2 Dataset description
2.3 Signal Decomposition

3 EEG functional connectivity methodologies
3.1 Synchronization Studies
3.1.1 Amplitude and Phase synchronization
3.1.2 Phase Locking Value (PLV)
3.2 Frequency dependent studies
3.2.1 Approximate (AE) and Transfer Entropy(TE)
3.2.2 Spectral Entropy

4 Phase Locking Value
4.1 Time-Frequency PLV spectrum
4.2 Identification of Differential Scalp Maps
4.3 Network Measures
4.3.1 Path length (PL) and clustering-coefficient explanation

5 Wavelet Spectral Entropy
5.1 Wavelet Spectral Entropy Time-Frequency Spectrum
5.2 Wavelet Spectral Entropy
5.3 Identification of Reactive band (RB) and Most Reactive Pairs (MRP)
5.4 Functional connectivity and Network Measures
5.5 Discussion

6 Conclusion and Future works

References

Preface

Functional mapping of the nervous system helps to identify the activity in the human brain during certain tasks. The difference between the two conditions helps you to understand the electrode nodes that cause a particular change in the neural state. Therefore, in this book, we will witness the functional difference between healthy people, Alzheimer's (AD) patients, multiple cognitive disorder (MCI) patients, and dementia patients. This gives the insight into the pair of electrodes that cause the progression of the disease from MCI to AD to dementia. This will help us in designing the biomarker in the future.

Ms. Padma Priya Vijayakumaran

Dr. Kalyana C. Veluvolu

Both authors are with the School of Electronics Engineering, Kyungpook National University, South Korea

List of Figures

Fig. 1 The overall workflow

Fig. 2 Steps in pre-processing

Fig. 3 Representation of major pre-processing steps involved in cleaning EEG

Fig. 4 Timeline progression of MCI, AD, dementia

Fig. 5 Electrode location(left) and experiment protocol(right)

Fig. 6 T-F PLV spectrum for Alzheimer’s disease(AD), multiple cognitive impairment (MCI), dementia(D) and normal(N) subjects

Fig. 7 The individual PLV connectivity maps and their differential maps between conditions

Fig. 8 Example for graph

Fig. 9 Path length and clustering coefficient for various thresholds

Fig. 10 The entropy time-frequency of healthy, MCI, AD and dementia subjects

Fig. 11 (A) dEntropy spectrum between healthy and MCI subjects. (B) dEntropy spectrum between healthy and AD subjects. (C) dEntropy spectrum between healthy and dementia subjects 33 Fig. 12 The WSE network and the differential functional connectivity map between all the conditions with constant threshold for all the conditions

Fig. 13 Path Length with different threshold for various conditions(left). clustering coefficients corresponding to all conditions(right)

List of Tables

Table 1 Data Description

Table 2 Average of Top5,10,15% Electrode pairs with maximum dPLV

1 Introduction

Electrophysiology is the study of electrical activities of bio cells and tissues and Electroencephalographic (EEG) is an electrophysiological monitoring technique used to record the electrical activity of the human brain. EEG is non-invasive and has high temporal resolution. Due to the emergence of various sectors like molecular biology, electrophysiology, and computational neuroscience the scientific study of neural system has been escalating from the second half of twentieth century. What do engineers have to do with brain? Neural engineering is one of the most significant and highly invested disciplinary supporting the neural research in the present society. It is a discipline within biomedical engineering that enhance the neural system with engineering techniques 1. Due to globalization and sedate life style, the present generation is facing a huge threat of getting affected by neural disorders. Neural disorders are any structural, biochemical or electrical abnormalities in neural system 2. The major neural disorders faced by the society includes Alzheimer’s disease (AD), Parkinson’s disease (PD), epilepsy, attention deficit hyperactive disorder (ADHD). In this study we will be analysing issues on Alzheimer’s disease. Why Alzheimer’s? Presently, 48.6 billion people are living with Dementia in this world, among them 70% are patients who have dementia due to AD. It is said that this rate will increase twice the present state by 2030. Governments are hugely investing on finding solution to this developing crisis. Scientists have been working to find an effective Bio marker to understand the process that takes place in the brain during this conversion period from AD to dementia. These bio markers will help the physicians to find the root cause of the development of AD. Understanding the activity inside the brain functional studies are indispensable, in 3,4 the functional connectivity, structural properties were analyzed with the help of graph theory. Graph theory helps in analysing the connectivity networks in the brain. The relatedness of the human neural system can be determined at the hand of different technical applications like Synchronization measures, cross-correlation, etc 5. In this thesis, we have dealt with one of the considerable techniques in practice, particularly wavelet spectral entropy (WSE) which tells you the randomness of the signals. In other words, it can be described as a measure of information or predictability of the signal. The considerable advantage in WSE is that it is independent of amplitude. For example, a basic sine wave is completely predictable irrespective of its strength of occurrence. In this paper, we will study the synchronous and the uniformity of the signals between two electrode pairs for every frequency and time instant using WSE. From the Time-Frequency maps, the reactive band (frequency band reacting to the condition) and the most reactive pairs to form the connectivity network are analyzed. It is clearly seen that WSE have its own assets and defects when it comes to form a differential functional connectivity network. Graph theory is widely applied for graphing the knowledge of the structure of all the social networks and in recent years it has been gaining interest in the field of neural networks 6. So, What is a graph? In mathematics graphs are pictorial representation of relation between two or more quantities. By joining two vertices, we can form a connection that provides the information of dependency or any other relatable quantity of the data used. In the perspective of human brain and EEG, EEG has electrodes placed on the scalp and thus we can consider them as vertices, based on the measure we use, and our objective, the connections can be made between those vertices. This will provide us with the information on the methodology we used and the response of brain to that methodology. In the literature, scientists have used graph theory for various applications in EEG like memory performance connectivity, cognitive activity analysis, functional connectivity that quantifies the cognitive activities and other neural properties 7. In this thesis we have dealt with the application of graph theory in finding the functional connectivity during the motor imagery (MI) event and also the functional difference between AD, dementia, multiple cognitive impairment (MCI) patients with the normal subjects. The Graphs can be drawn by defining the connectivity through the methodologies used in the analysis. Techniques related Synchronization measures, correlations etc, provide the similarity or the synchronization between the electrodes in human brain.

The basic work flow used is explained in Fig. 1. Once the EEG is collected from the subjects the first technique applied is pre-processing of the signal. The various steps involved in pre-processing are shown in Fig. 2 and Fig. 3.

Abbildung in dieser Leseprobe nicht enthalten

The general analysis proceeds to application of data mining methods (in our scenario) or any other technique to analyse the information present in the data. Secondly, the data in time domain are transformed into frequency domain, as we are dealing with phase and frequency of the signal. This transformation is carried out using complex morlet wavelet transform. The explanation on wavelet transform is explained in chapter 2.

Once the signal decomposition is completed, the techniques like phase locking value (PLV), wavelet spectral entropy (WSE) etc.. are applied to extract the significant information from time-frequency spectrums. Threshold is set according to the requirement and the electrode pairs reacting to that particular protocol are considered to be connected. Similarly, differential network plots between the different conditions are found by finding the corresponding differential matrix.

In this thesis, the analysis goes as follows, in the following chapter we explain the major neural disorder which is considered for analysis. In chapter 3, the various information theory methodology used to analyse the signals are explained with their applications in neuroscience.In chapter 4, the first measure namely phase locking value (PLV) is analysed with AD data-set and the results are described. In chapter 5, another method namely, wavelet spectral entropy (WSE) is used to evaluate the differential network between AD, MCI, dementia and healthy subjects to design a biomarker. Finally, the thesis is concluded and discussed on future work is discussed.

2 Neural Disorders, Dataset Description and Preliminary Signal Processing

With Increasing influence of technology and mentally draining jobs, our current generation is undoubtedly subjected to high levels of stress and risk of neural disorders. Unhealthy life style, nuclear family, less physical activity, less human interactions etc also places a huge role in causing neural diseases. Few of the major diseases include Alzheimer’s Disease (AD) and Attention Deficit/hyperactive Disorder (ADHD), where AD affects the elderly humans and ADHD being a major threat to children below their adolescent age.

2.1 Alzheimer’s Disease

Alzheimer’s Disease (AD) is a prevailing significant hazard to the present society 6, 50 million people are living with dementia in the world among which 70% are dementia caused by AD 10 and it is said that 10 million people get diagnosed with dementia every year. AD majorly affects the elderly society in the civilization. One of the concerns with AD is it is irreversible, they tend to progress with time and eventually the patient will have to rely completely on a personal care-taker to do their everyday chores. The symptoms of highly submissive and are mostly misinterpreted for the signs of aging, as its noticeable only over the age of 60 years. As a consequence, the patients will have problems with remembering all the past activities initially, their recalling ability will be damaged due to neural degeneration that causes the inability to perform cognitive tasks 11. AD has been declared as sixth reason that leads to death in United States of America but it is estimated that this ranking will rise up to third after heart diseases and cancer being the major cause of death of elder people. Dr. Alois Alzheimer was the one who discovered the unusual presence of amyloid plaques and fiber bundles 12 in the neural system, that’s first discovery of AD . The disease then was named after him. Those amyloid plaques and the fiber tangles are considered to be a significant reason for the cause of AD. The other major observation is that the neurons connectivity is highly reduced or loss of connectivity and complexity between neurons. It is evident in functional connectivity mapping of AD patients. In the further chapters we will be able to see the difference between all the functional connectivity between the different pathologies in comparison namely MCI, dementia, and healthy subjects. The progression of the disease is shown in Fig. 4

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Timeline progression of MCI, AD, dementia.

2.1.1 Attention deficit/hyperactive disorder (ADHD)

Attention deficit/hyperactive disorder (ADHD) is a psychiatric anarchy of the brain. It is diagnosed when the subject shows notable abstractive, hyperactive and impetuous behaviors that hinders the neural development 13.

- Abstraction refers to a state where one is not being able to focus in a particular task for a long period of time. They will not be able to endure anything for a longer course of time.
- Hyperactive This refers to a state where the individual will always be engaged in a work. This person can never sit in a place without doing anything not even for a short period. They tend to move, jump and act inappropriate to the situation.
- Impetuous This can be explained as hastily response towards any situation. The subjects with ADHD won’t think twice, they act instantly to everything. They decide with the first thought that comes to their mind and act in haste.

As there is a huge demand for quantitative studies required in the field of Neural disorders, scientists have been trying to explore more on similar Neural degenerative and other similar Brain malfunctions. Along with AD and ADHD. Neurological disorders like epilepsy, seizures, stroke are also a huge concern for the doctors because detection, quantitative analysis, recording the progression, mathematical proofs are all a huge validations doctors need in order to formulate a treatment for every patient and moreover all the brain related studies are subject specific thus both scientific and mathematical explanations are focused in present day scenario.

2.2 Dataset description

The dataset comprises of 31 subjects(17 female, 13 male, mean age of 75.1 and standard deviation of 7.1) in which 14 belong to AD, 6 to MCI, 5 to dementia and 5 to healthy condition, they were all recorded at Neurology & Neurosurgery Center, Kyungpook National University Chilgok Hospital, Daegu, Republic of Korea. The classification of patients were made by considering their mini mental state exam (MMSE) score, CDR scores, visible symptoms, physician’s consultation, MRI, PET etc. The 4 groups of subjects had the following MMSE scores (out of 30), the AD subjects’ MMSE scores were in the range of 5 - 20, dementia patients’ scores were between 11 — 16, MCI patients’ scores ranged 2426 and healthy subjects had MMSE scores > 27. The complete EEG data which were recorded contains 23 Ag/AgCl electrodes. The signals were band pass filtered between 0.5 - 50 Hz in which 1 - 45 Hz were considered for analysis and that includes δ (1 - 4 Hz), θ (4 — 7 Hz), α (8 - 13 Hz), β (15 - 30 Hz), and low γ (30 - 45 Hz) frequency bands 3. The data comprises of 20 minutes EEG collected from healthy subjects, MCI, AD and dementia patients in which 10 minutes is of eyes closure (in rest state) and 10 minutes is of eyes open condition. During the experiment period, the subjects were asked to reduce or avoid any movements to evade the artifacts that can occur due to motor activity and excessive eye blinks. These signals were sampled with a sampling frequency of 500 Hz and then later downsampled to 256 Hz for offline analysis. The ECG electrode signal, A1, A2 (reference electrodes), Human Pulse Oximetry electrode (SpO2), EOG electrode signals were neglected during analysis. Electrode placements and the experiment protocols are shown in Fig. 1.

Abbildung in dieser Leseprobe nicht enthalten

The detailed description of the data used in Table. 1 For further analysis in this research only data corresponding to eyes open state is considered due to instability of AD patients during eyes closed rest state. As you can see in the table, the sampling frequency and the number of channels of the data highly differs from subject to subject. This was re sampled to 256Hz and 23 common channels which where present in all the subject data.The electrodes channels considered are Fp1, Fp2, F7, F3, Fz, F4, F8, F9, F10, T9, T10, T7, C3, Cz, T8, P7, P3, Pz, P4, P8, O1, O2.

[...]

Excerpt out of 54 pages

Details

Title
Brain Mapping for Dementia Disorder
Grade
2.0
Authors
Year
2020
Pages
54
Catalog Number
V957909
ISBN (eBook)
9783346312631
ISBN (Book)
9783346312648
Language
English
Keywords
brain, mapping, dementia, disorder
Quote paper
Kalyana Veluvolu (Author)Padma Priya Vijayakumaran (Author), 2020, Brain Mapping for Dementia Disorder, Munich, GRIN Verlag, https://www.grin.com/document/957909

Comments

  • No comments yet.
Look inside the ebook
Title: Brain Mapping for Dementia Disorder



Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free