M. Aswatha Kumar

2012 Special Issue: Identification of brain regions responsible for Alzheimer's disease using a Self-adaptive Resource Allocation Network

In this paper, we present a novel approach for the identification of brain regions responsible for Alzheimers disease using the Magnetic Resonance (MR) images. The approach incorporates the recently developed Self-adaptive Resource Allocation Network (SRAN) for Alzheimers disease classification using voxel-based morphometric features of MR images. SRAN classifier uses a sequential learning algorithm, employing self-adaptive thresholds to select the appropriate training samples and discard redundant samples to prevent over-training. These selected training samples are then used to evolve the network architecture efficiently. Since, the number of features extracted from the MR images is large, a feature selection scheme (to reduce the number of features needed) using an Integer-Coded Genetic Algorithm (ICGA) in conjunction with the SRAN classifier (referred to here as the ICGA-SRAN classifier) have been developed. In this study, different healthy/Alzheimers disease patients MR images from the Open Access Series of Imaging Studies data set have been used for the performance evaluation of the proposed ICGA-SRAN classifier. We have also compared the results of the ICGA-SRAN classifier with the well-known Support Vector Machine (SVM) and Extreme Learning Machine (ELM) classifiers. The study results clearly show that the ICGA-SRAN classifier produces a better generalization performance with a smaller number of features, lower misclassification rate and a compact network. The ICGA-SRAN selected features clearly indicate that the variations in the gray matter volume in the parahippocampal gyrus and amygdala brain regions may be good indicators of the onset of Alzheimers disease in normal persons.

Symmetric Key Image Encryption Scheme with Key Sequences Derived from Random Sequence of Cyclic Elliptic Curve Points

In this paper, cyclic elliptic curves of the form y^2+xy=x^3+ax^2+b,a,b  GF(2(superscript m)) with order M is considered in the design of a Symmetric Key Image Encryption Scheme with Key Sequence derived from random sequence of cyclic elliptic Curve points. P with co-ordinates (xP, yP) which satisfy the elliptic curve equation is called a point on elliptic curve. The order M is the total number of points (x, y) along with x=∞, y=∞ which satisfy the elliptic curve equation. Least integer N for which NP is equal to point at infinity O is called order of point P. Then P, 2P,…. (N-1) P are distinct points on elliptic curve. In case of cyclic elliptic Curve there exists a point P having the same order N as elliptic curve order M. A finite field GF (p) (p≥N) is considered. Random sequence {k(subscript i)} of integers is generated using Linear Feedback Shift Register (LFSR) over GF (p) for maximum period. Such sequences are called maximal length sequences and their properties are well established. Every element in sequence {k(subscript i)} is mapped to k(subscript i) P which is a point on cyclic elliptic Curve with co-ordinates say (x(subscript i), y(subscript i)). The sequence {k(subscript i)} is a random sequence of elliptic curve points. From the sequence (x(subscript i), y(subscript i)) several binary and non-binary sequences are derived. These sequences find applications in Stream Cipher Systems. Two encryption algorithms-Additive Cipher and Affine Cipher are considered. Results of Image Encryption for a medical image is discussed in this paper. Here, cyclic elliptic Curve over GF(2^8) is chosen for analysis.

Alzheimer's disease detection using a Self-adaptive Resource Allocation Network classifier

This paper presents a new approach using Voxel-Based Morphometry (VBM) detected features with a Self-adaptive Resource Allocation Network (SRAN) classifier for the detection of Alzheimers Disease (AD) from Magnetic Resonance Imaging (MRI) scans. For feature reduction, Principal Component Analysis (PCA) has been performed on the morphometric features obtained from the VBM analysis and these reduced features are then used as input to the SRAN classifier. In our study, the MRI volumes of 30 ‘mild AD to moderate AD’ patients and 30 normal persons from the well-known Open Access Series of Imaging Studies (OASIS) data set have been used. The results indicate that the SRAN classifier produces a mean testing efficiency of 91.18% with only 20 PCA reduced features whereas, the Support Vector Machine (SVM) produces a mean testing efficiency of 90.57% using 45 PCA reduced features. Also, the results show that the SRAN classifier avoids over-training by minimizing the number of samples used for training and provides a better generalization performance compared to the SVM classifier. The study clearly indicates that our proposed approach of PCA-SRAN classifier performs accurate classification of AD subjects using reduced morphometric features.

Random Binary and Non-Binary Sequences Derived from Random Sequence of Points on Cyclic Elliptic Curve Over Finite Field GF(2m) and Their Properties

In this paper, Cyclic Elliptic Curves of the form y 2 + xy = x 3 + ax 2 + b,a,b ∈ GF(2 m ) with order M is considered. A finite field GF(p) (p ≥ N, where N is the order of point P) is considered. Random sequence {k i } of integers is generated using Linear Feedback Shift Register (LFSR) over GF(p) for maximum period. Every element in sequence {k i } is mapped to k i P which is a point on Cyclic Elliptic Curve with co-ordinates say (x i , y i ). The sequence {k i P} is a random sequence of elliptic curve points. From the sequence (x i , y i ) several binary and non-binary sequences are derived and their randomness properties are investigated. The results are discussed. It is found that these sequences pass FIPS-140, NIST tests and exhibit good Hamming Correlation properties. These sequences find applications in Stream Cipher Systems. Here, Cyclic Elliptic Curve over GF(28) is chosen for analysis.

Generation of pseudorandom sequence over elliptic curve group and their properties

Abstract In this paper, the application of Elliptic Curves in the generation of pseudorandom sequences is discussed. An Elliptic Curve of the form y 2+xy=x 3+ax 2+b, over GF(28) is considered. A base point P of large order N, of elliptic curve has been found. Using Linear Feedback Shift Register (LFSR) for generating sequence of integers {k i } modulo p, a prime, where p≥N, from which random key sequence {k i P} is obtained. The details of randomness tests [9] performed on the binary form of the sequence {k i P} of Elliptic Curve points are discussed for a sample example. The generated sequences have passed most of the FIPS 140–2 statistical tests. Hamming Correlation of random sequences generated is determined and results are discussed. Such sequences find applications as key sequences in the Stream Cipher Systems. Some of the results obtained in such applications are also given in this paper.

Alzheimer’s Disease Detection Using Minimal Morphometric Features with an Extreme Learning Machine Classifier

In this paper, we present an accurate method of detection of Alzheimer’s disease using a minimal number of voxel-based morphometry features obtained from the brain MRI scans. The problem of early detection of AD is formulated as a binary classification problem and solved using an extreme learning machine classifier. The functional relationship between the voxel-based morphometry features extracted from magnetic resonance images and Alzheimer’s disease is approximated closely using the extreme learning machine classifier. Since, the extreme learning machine is computationally efficient and provides a better generalization ability, Principal Component Analysis along with the Extreme Learning Machine classifier (referred to here as the PCA-ELM classifier) is used to select the minimal set of morphometric features from the brain MRI images for Alzheimer’s disease detection. Performance of the PCA-ELM classifier is evaluated using the Open Access Series of Imaging Studies (OASIS) data set. The results are also compared with the well-known support vector machine classifier. The study results clearly show that the PCA-ELM classifier produces a better generalization performance with a minimal set of features.

Project Based Learning in Higher Education with ICT: Designing and Tutoring Digital Design Course at M S R I T, Bangalore

This paper presents an approach to develop digital design curricula using modified Bloom’s Taxonomy for making it more appealing to the students. The proposed approach uses the Project Based Learning strategy for increasing the attractiveness of the course. The case study also explores the use of ICT for effective teaching and learning. The developed curriculum has been evaluated successfully for the last three academic years. The students have shown increased interest in the digital design course, have acquired new skills and obtained good academic results. An important observation during the conduction of the course was that all the students have developed innovative digital systems, exhibited team work and have made a poster presentation during their second year of engineering undergraduate course.

A Neural Network Based Solution to Color Image Restoration Problem

In this paper, the problem of color image restoration using a neural network learning approach is addressed. Instead of explicitly specifying the local regularization parameter values, we modify the neural network weights, which are considered as the regularization parameters. These are modified through the supply of appropriate training examples. The desired response of the network is in the form of estimated value for the current pixel. This estimate is used to modify the network weights such that the restored value produced by the network for a pixel is closer to this desired response. In this way, once the neural network is trained, images can be restored without having prior information about the model of noise/blurring with which the image is corrupted.