Principal component analysis of the spectrogram of the speech signal: Interpretation and application to dysarthric speech

Long-Term Averaged Spectrum Descriptors of Dysarthria in Patients With Parkinson's Disease Treated With Subthalamic Nucleus Deep Brain Stimulation

PURPOSE

This study aimed to evaluate whether long-term averaged spectrum (LTAS) descriptors for reading and monologue are suitable to detect worsening of dysarthria in patients with Parkinson's disease (PD) treated with subthalamic nucleus deep brain stimulation (STN-DBS) with potential effect of ON and OFF stimulation conditions and types of connected speech.

METHOD

Four spectral moments based on LTAS were computed for monologue and reading passage collected from 23 individuals with PD treated with bilateral STN-DBS and 23 age- and gender-matched healthy controls. Speech performance of patients with PD was compared in ON and OFF STN-DBS conditions.

RESULTS

All LTAS spectral moments including mean, standard deviation, skewness, and kurtosis across both monologue and reading passage were able to significantly distinguish between patients with PD in both stimulation conditions and control speakers. The spectral mean was the only LTAS measure sensitive to capture better speech performance in STN-DBS ON, as compared to the STN-DBS OFF stimulation condition (p < .05). Standardized reading passage was more sensitive compared to monologue in detecting dysarthria severity via LTAS descriptors with an area under the curve of up to 0.92 obtained between PD and control groups.

CONCLUSIONS

Our findings confirmed that LTAS is a suitable approach to objectively describe changes in speech impairment severity due to STN-DBS therapy in patients with PD. We envisage these results as an important step toward a continuum development of technological solutions for the automated assessment of stimulation-induced dysarthria.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.21644798.

A New Fractional-Order Chaotic System with Its Analysis, Synchronization, and Circuit Realization for Secure Communication Applications

This article presents a novel four-dimensional autonomous fractional-order chaotic system (FOCS) with multi-nonlinearity terms. Several dynamics, such as the chaotic attractors, equilibrium points, fractal dimension, Lyapunov exponent, and bifurcation diagrams of this new FOCS, are studied analytically and numerically. Adaptive control laws are derived based on Lyapunov theory to achieve chaos synchronization between two identical new FOCSs with an uncertain parameter. For these two identical FOCSs, one represents the master and the other is the slave. The uncertain parameter in the slave side was estimated corresponding to the equivalent master parameter. Next, this FOCS and its synchronization were realized by a feasible electronic circuit and tested using Multisim software. In addition, a microcontroller (Arduino Due) was used to implement the suggested system and the developed synchronization technique to demonstrate its digital applicability in real-world applications. Furthermore, based on the developed synchronization mechanism, a secure communication scheme was constructed. Finally, the security analysis metric tests were investigated through histograms and spectrograms analysis to confirm the security strength of the employed communication system. Numerical simulations demonstrate the validity and possibility of using this new FOCS in high-level security communication systems. Furthermore, the secure communication system is highly resistant to pirate attacks. A good agreement between simulation and experimental results is obtained, showing that the new FOCS can be used in real-world applications.

BrainOS: A Novel Artificial Brain-Alike Automatic Machine Learning Framework

Human intelligence is constituted by a multitude of cognitive functions activated either directly or indirectly by external stimuli of various kinds. Computational approaches to the cognitive sciences and to neuroscience are partly premised on the idea that computational simulations of such cognitive functions and brain operations suspected to correspond to them can help to further uncover knowledge about those functions and operations, specifically, how they might work together. These approaches are also partly premised on the idea that empirical neuroscience research, whether following on from such a simulation (as indeed simulation and empirical research are complementary) or otherwise, could help us build better artificially intelligent systems. This is based on the assumption that principles by which the brain seemingly operate, to the extent that it can be understood as computational, should at least be tested as principles for the operation of artificial systems. This paper explores some of the principles of the brain that seem to be responsible for its autonomous, problem-adaptive nature. The brain operating system (BrainOS) explicated here is an introduction to ongoing work aiming to create a robust, integrated model, combining the connectionist paradigm underlying neural networks and the symbolic paradigm underlying much else of AI. BrainOS is an automatic approach that selects the most appropriate model based on the (a) input at hand, (b) prior experience (a history of results of prior problem solving attempts), and (c) world knowledge (represented in the symbolic way and used as a means to explain its approach). It is able to accept diverse and mixed input data types, process histories and objectives, extract knowledge and infer a situational context. BrainOS is designed to be efficient through its ability to not only choose the most suitable learning model but to effectively calibrate it based on the task at hand.