Isabella Palamara

Multivariate Multi-Scale Permutation Entropy for Complexity Analysis of Alzheimer’s Disease EEG

Abstract: An original multivariate multi-scale methodology for assessing the complexity of physiological signals is proposed. The technique is able to incorporate the simultaneous analysis of multi-channel data as a unique block within a multi-scale framework. The basic complexity measure is done by using Permutation Entropy, a methodology for time series processing based on ordinal analysis. Permutation Entropy is conceptually simple, structurally robust to noise and artifacts, computationally very fast, which is relevant for designing portable diagnostics. Since time series derived from biological systems show structures on multiple spatial-temporal scales, the proposed technique can be useful for other types of biomedical signal analysis. In this work, the possibility of distinguish among the brain states related to Alzheimer’s disease patients and Mild Cognitive Impaired subjects from normal healthy elderly is checked on a real, although quite limited, experimental database. Keywords:

Entropic Measures of EEG Complexity in Alzheimer's Disease Through a Multivariate Multiscale Approach

Alzheimers disease (AD) impact is rapidly growing in western countries. The unavoidable progression of the disease, call for reliable ways to diagnose the AD in its early stages. Recently, it has been shown that the electroencephalography (EEG) complexity analysis could be used to predict the conversion from mild cognitive impairment (MCI) to AD. Despite the EEG analysis does not achieve yet the required clinical performance in terms of both sensitivity and specificity to be accepted as a clinically reliable technique of screening, the researchers count on the easiness and the non-invasiveness of the EEG measuring system. The aim of this paper is to analyze the efficacy of entropic complexity measures as a possible bio-marker to distinguish among the brain states related to the AD patients and MCI subjects from normal healthy elderly. The research is carried out on an experimental database. Three different emerging measures of complexity are compared, namely, permutation entropy, sample entropy, and Lempel-Ziv complexity. Because time series derived from biological systems show structures on multiple spatial-temporal scales and there exists a significant inter-channel correlation among the EEG channels, a multiscale multivariate approach is also implemented. Limited to the analyzed data, the results show that the severity of the AD reflects in the EEG dynamic complexity leaving the hope of early diagnosis based on simple EEG.

EEG Complexity Modifications and Altered Compressibility in Mild Cognitive Impairment and Alzheimer’s Disease

The objective of this work is to respond to the question: can quantitative electroencephalography (EEG) distinguish among Alzheimer’s Disease (AD) patients, mild cognitive impaired (MCI) subjects and elderly healthy controls? In other words, are there nonlinear indexes extracted from raw EEG data that are able to manifest the background difference among EEG? The response we give here is that a synthetic index of entropic complexity (Permutation Entropy, PE) as well as a measure of compressibility of the EEG can be used to discriminate among classes of subjects. An experimental database has been analyzed to make these measurements and the results we achieved are encouraging also in terms of disease evolution. Indeed, it is clearly shown that the condition of MCI has intermediate properties with respect to the analyzed markers: thus, these markers could in principle be used to evaluate the probability of transition from MCI to mild AD.

Optical Performance Evaluation of Oil Spill Detection Methods: Thickness and Extent

The release of petroleum liquids in water, such as marine, riverine, and lacustrine basins, is a matter of concern that shoves authorities and experts to adopt technical approaches for preventing damages, monitoring oil content in water and cleaning up environmental aqueous matrices. In this paper, a dedicated system for oil spill detection is presented. The proposed system consists of an optical fiber sensor and an image processing unit useful to steer and optimize the measurement process. An optical fiber-based antenna sensor is used to detect the oil concentration in water. The sensor consists of two slanted optical probes acting as transmitter and receiver, respectively. Both probes are completely immersed into water being analyzed. The sensing approach is based on the measurement of the light coupling level affected by the reflectivity of the oil layer floating on the water surface. The experimental measurement of different types of oil is performed to assess the sensitivity of the developed system. Special attention is put on the image postelaboration useful to derive the characteristics of the oil distribution on the water surface. In this respect, two different image processing techniques are considered: the first one is based on a suitable energy-minimizing spline fitting procedure subject to external constraint forces, whereas a judicious use of the Hough transform is made in the second one.

SVM classification of epileptic EEG recordings through multiscale permutation entropy

Electroencephalogram (EEG) is a non-invasive diagnostic tool in clinical neurophysiology, especially with respect to epilepsy. The epileptic status is characterized by reduced complexity. New markers, based on nonlinear dynamics, like Permutation Entropy (PE) have been developed to measure EEG complexity. In this paper, Multiscale Permutation Entropy (MPE) complexity measure is proposed as a potentially useful framework for detecting epileptic events in EEG data and to distinguish healthy controls from patients. The achieved results show that: 1) MPE is able to discriminate between the two categories; 2) the use of multiple scales may substantially improve the specificity of the diagnosis. This is shown through an SVM-based classification network with three different kernels. The use of the SVM approach is also useful to infer clues about the extracted features.

Fuzzy similarity measures for detection and classification of defects in CFRP

The systematic use of nondestructive testing assumes a remarkable importance where on-line manufacturing quality control is associated with the maintenance of complex equipment. For this reason, nondestructive testing and evaluation (NDT/NDE), together with accuracy and precision of measurements of the specimen, results as a strategic activity in many fields of industrial and civil interest. It is well known that nondestructive research methodologies are able to provide information on the state of a manufacturing process without compromising its integrity and functionality. Moreover, exploitation of algorithms with a low computational complexity for detecting the integrity of a specimen plays a crucial role in real-time work. In such a context, the production of carbon fiber resin epoxy (CFRP) is a complex process that is not free from defects and faults that could compromise the integrity of the manufactured specimen. Ultrasonic tests provide an effective contribution in identifying the presence of a defect. In this work, a fuzzy similarity approach is proposed with the goal of localizing and classifying defects in CFRP in terms of a sort of distance among signals (measure of ultrasonic echoes). A field-programmable gate array (FPGA)-based board will be also presented which implements the described algorithms on a hardware device. The good performance of the detection and classification achieved assures the comparability of the results with the results obtained using heuristic techniques with a higher computational load.

On the Use of Empirical Mode Decomposition (EMD) for Alzheimer’s Disease Diagnosis

Alzheimer’s Disease (AD) is considered one of the most common form of dementia; it involves a progressive decline in cognitive function because of pathological modifications or damage of the brain. One of the major challenges is to develop tools for early diagnosis and disease progression. Electroencephalogram represents potentially a noninvasive and relatively non-expensive approach for screening of dementia and AD. It provides a method to objectively quantify the cortical activation patterns but it is usually considered insensitive in the early AD. This study introduces a novel method where electroencephalographic recordings (EEG) are subjected to Empirical Mode Decomposition (EMD), which decomposes a signal into components known as Intrinsic Mode Functions (IMFs). The results, suggest that, the IMFs may be used to determine the particular frequency bandwidths in which specific phenomena occur.

Innovative Fuzzy Techniques for Characterizing Defects in Ultrasonic Nondestructive Evaluation

Classification of defects in ultrasonic nondestructive testing (NDT)/nondestructive evaluation (NDE) has a role of primary importance in all those applications in which the knowledge of the typology of defect is crucial for the manufact destination. In such a context, the necessity to have efficient investigation instruments for a correct classification analysis emerges clearly. A defect, even when invisible to the naked eye, can be revealed as the cause that reduces the similarity of a measured signal with respect to a reference. Considering the fuzziness intrinsic in the signals, the reliance on fuzzy techniques to evaluate similarity appears desirable. Two fundamental achievement of research in such field, which both derive from the fuzzy thinking and which share common traits, are computing with words (CW) and the concept of fuzzy similarity (FS) . In CW, a word is considered as a label of a fuzzy set of points clustered by similarity (granule) which lead to a particular formulation of a bank of fuzzy rules structured per classes. FS is an evaluation index of similarity among entities (for example signals) particularly useful for the formation of specific classes. Both approaches are based on computational linguistics (for example, descriptive formalism in natural language). This chapter is conceptually divided into two parts: the first one is dedicated to the development of detection and classification techniques of defectiveness for ultrasonic NDE by means of CW, while the second one, with the same purpose, proposes an approach based on FS.

Fuzzy similarities for the optimization of the operational parameters related to ultrasonic phased array probes for industrial application

The systematic use of non-destructive testing is important both for on-line quality control during the production process and the maintenance of complex equipment especially in industrial applications. In such a context, the ultrasound imaging techniques cover an important role above all in those cases in which it is requested a great penetration depth. Starting from an experimental database (affected by uncertainties due to sampling and/or noise, so of clear fuzzy nature) , the authors propose a new approach aimed at the optimization of the operational parameters of ultrasonic phased array probes. the problem of optimization is translated in an equivalent problem of classification through a fuzzy similarity approach: different images produced by different probes are compared. The quality of the proposed procedure has been assessed comparing the obtained results also with SOM- type technique in which the parameters of similarity among neurons has been considered of fuzzy nature.

Standard Soft Computing Techniques for Characterization of Defects in Nondestructive Evaluation

Nondestructive evaluation (NDE)/nondestructive testing (NDT) of industrial manufactured items holds a strategic importance both for the quality assessment of the productive process and for taking monitoring actions during the relative life cycle. Since the quality assessment passes through the characterization of the defects that can be generated, it is imperative to exploit inspection techniques able to produce signals characterizing the defects themselves. Current inspection techniques do not provide, for several reasons, signals free from errors, inaccuracies and imprecision, so, in the information-processing step, it is necessary to face the problem by approaches capable of taking into account the inherent vagueness . While the form reconstruction of a defect is still an open problem, its localization and classification has been carried out with excellent results by the scientific community with the development of efficient and accurate methodologies also in terms of vagueness management . Considering the technological transfer point of view, the approaches developed so far are burdened by a less than desirable computational complexity that translates into expensive hardware requirements. For this reason, it is necessary to elaborate alternative methodologies capable of combining high-quality results and low-computational complexity. Specifically, among the many possible computing techniques, attention is given to the soft techniques, and in particular, to the fuzzy logic (FL). The latter, generalizing dichotomic logic and taking into account the vagueness of signals, can deliver results comparable as a whole to those obtainable by more sophisticated techniques, but with a reduced computational charge. Moreover, the formalization in terms of natural language (NL) leads to the structuring of systems managed by legible linguistic rules even by nonexperts in the field and, at the same time, easily revisable by the expert. The present chapter is completely dedicated to introduce the reader to the basic principles of the logic of fuzzy inference system (FIS), applied functionally within NDE/NDT, presenting an example of study of the inverse problem.