Alice N. Cheeran

Multitaper perceptual linear prediction features of voice samples to discriminate healthy persons from early stage Parkinson diseased persons

The performance of multitaper perceptual linear prediction (PLP) features of speech samples to discriminate healthy and early stage Parkinson diseased subjects is investigated in this paper. The PLP features are conventionally obtained by computing the power spectrum using a single tapered Hamming window. This estimated spectrum exhibits large variance which can be reduced by computing the weighted average of power spectra obtained using a set of tapered windows, leading to multitaper spectral estimation. In this investigation, two multitaper techniques namely Sine wave taper and Thomson multitaper along with the conventional single taper windowing are investigated. Artificial Neural network is then used to classify the PLP features extracted by applying the three types of window tapers on the speech signals of healthy and early stage Parkinson affected people and their respective performances are compared. The results show more accuracy using the multitaper techniques when compared with the conventional single taper technique. It is seen that the accuracy obtained using Sine wave tapers as well as Thomson multitaper is maximum for five tapers. An improvement in the recognition accuracy by 7.5% using the Sine tapers and by 6.9% using the Thomson tapers is obtained when compared with the conventional method. An improvement in other performance measures like Equal error rate, False positive rate, False negative rate, Sensitivity and Specificity is also observed in the multitaper techniques.

Thomson Multitaper MFCC and PLP voice features for early detection of Parkinson disease

Abstract In this paper, MFCC and PLP voice features extracted using Single Taper Smooth (STS) window and Thomson Multitaper (TMT) windowing technique together with a neural network classifier is used in the classification of Healthy people from early stage Parkinson diseased patients and a performance comparison of the two techniques is reported. Parkinson disease in their early stages, not only affects the muscular movements of the human body but also influences the articulatory process of the speech production mechanism. This signifies change in the shape of the vocal tract which manifests itself in the short time power spectrum. The MFCC and PLP features used in this investigation, which represent the vocal tract parameters are derived from the short time spectrum. It is therefore crucial to estimate this short time power spectrum accurately. Generally, the short time speech power spectrum is estimated using STS window. But this power spectrum computed manifests large variance in the spectral estimates. Hence a variance reduced power spectrum is attained by computing the weighted average of the short time speech spectra obtained using a set of TMT windows. This spectrum is then used to compute the PLP and MFCC features. In this paper, extraction of both these voice features using STS window as well as TMT technique with three different weights namely Uniform, Eigen value (EV) and Adaptive weights is implemented using the speech samples of healthy and Parkinson diseased individuals. The experiment was carried out for several Thomson tapers ranging from 1 to 12 and the optimal number of tapers needed for the application and dataset is reported. A comparative performance analysis of the techniques implemented using both MFCC and PLP as features is then carried out in terms of classification accuracy, Equal Error Rate, sensitivity, selectivity and F1 score for the optimal taper value. The results obtained show that in comparison with the STS window a maximum improvement in the classification accuracy was obtained to be 6.6% for nine tapers, adaptive weights using MFCC as features and 6.9% for five tapers, EV weights using PLP as features for experimental dataset 1 and 6.0% using MFCC and 6.4% using PLP for experimental dataset 2. A performance improvement in other measures for the optimal taper value is also observed and reported for experimental dataset 1.

Statistical comparison of Jitter and Shimmer voice features for healthy and Parkinson affected persons

Parkinson disease (PD) is a neurological diseaseoccurring because of the loss or impairment of the cells that produce dopamine present in the substantia nigra of the brain. This insufficiency affects the movement of the muscles and hence people with Parkinson develop speech impairments related to phonation subsystem which influences the voice features, Jitter and Shimmer. The analysis of these parameters is therefore helpful in diagnosing the disease and also in predicting the severity of the disease. In this paper, statistical analysis of several types of Jitter and Shimmer parameters is carried out on voice samples taken from Parkinson affected and healthy subjects. The various analyses carried out are the descriptive statistics, box plots, correlation analysis, Shapiro-Wilk Test and Mann-Whitney U test. The analysis performed shows a significant difference in the Jitter and Shimmer parameters between the two categories of voice samples.

Design of a quadrature hybrid waveguide coupler at high frequency for high power applications

In recent times, research in the field of nuclear fusion systems has been on the rise. Such systems play a vital role in the study of plasma physics and for possible applications towards the generation of electricity. Designing the components for such systems is a challenge owing to the demand for operation in high-frequency region (in the range of GHz) at extremely high powers. This paper presents a structure and methodology for designing a quadrature (90°) hybrid coupler operating at 3.7 GHz and 5 GHz frequencies. Designs have been simulated in COMSOL Multi physics software and RF heating results along with scattering parameters are presented in this paper.

Cubic spline approximation of bilinear frequency unwarping

Nonlinear frequency warping is of recent interest in speech processing studies as due to its ability to extract speaker specific features. In particular, all pass Bilinear Transformation (BLT) is highly recommended tool in realizing such desired frequency-warping. Proposed work suggests a generic approach to unwarp formerly unmanageable second order BLT frequency mappings. Various spline approximation techniques have been analyzed to fit the inverse warping functions. Experiments have been performed to inverts first and second order warping approximations of state of the art Mel scale and bark scale. This method has been employed to derive inverse triangular filter bank for arbitrary frequency warping.

Design data for quick development of Folded E plane Tee

Most of the microwave communication systems have requirement of power dividers that are essential for power splitting and combining operations. This paper presents a structure and methodology for designing a rectangular waveguide Folded E plane Tee. The structure proposed has the advantage of less area consumption as compared to a conventional waveguide Tee. The paper also presents design equations using which one can design a Folded E plane Tee at any desired frequency. The designs thus obtained at some random frequencies from the equations have been simulated in COMSOL Multiphysics and the scattering parameters obtained have been presented.

Glottal pathology discrimination using ANN and SVM

Use of modern technological advances in real-time biomedical analysis is very crucial. Current work focuses on glottal pathology discrimination based on non-invasive speech analysis techniques. Primary set back in developing such method is irregular performance depreciation of several state of the art acoustic features. To excuse such problems, we have used glottal to noise excitation ratio, which predicts the breathiness quotient of the speech signal and is supported by characteristic mean pitch value. To build a judicial model, we have used Artificial Neural Network (ANN) and Support Vector Machine (SVM). Categorization performance is compared using well known parameters like true positive rate, true negative rate and accuracy. Results of the analysis show slightly favored performance for SVM based decisive system.

A multiphysics investigation of RF dry loads

Microwave absorbers/loads find applications in many microwave systems where they are utilised in testing or protection circuits. In many waveguide based systems, often lossy dielectrics are used as absorbers. Efficient thermal management of the heat generated in the dielectric is crucial for the functioning of the load. As such the design of a load is a multiphysics problem involving RF, thermal and in some cases fluid flow analysis. In this paper, different structures of low power loads are analysed for their RF and thermal characteristics and behaviour. These loads are designed for 3.7 GHz and simulated in COMSOL Multiphysics.