Voice for Health: The Use of Vocal Biomarkers from Research to Clinical Practice

in Digital Biomarkers (2021), 5(1), 78-88

Diseases can affect organs such as the heart, lungs, brain, muscles, or vocal folds, which can then alter an individual’s voice. Therefore, voice analysis using artificial intelligence opens new ... [more ▼]

Diseases can affect organs such as the heart, lungs, brain, muscles, or vocal folds, which can then alter an individual’s voice. Therefore, voice analysis using artificial intelligence opens new opportunities for healthcare. From using vocal biomarkers for diagnosis, risk prediction, and remote monitoring of various clinical outcomes and symptoms, we offer in this review an overview of the various applications of voice for health-related purposes. We discuss the potential of this rapidly evolving environment from a research, patient, and clinical perspective. We also discuss the key challenges to overcome in the near future for a substantial and efficient use of voice in healthcare. [less ▲]

Design of a 2-Bit Neural Network Quantizer for Laplacian Source

in Entropy (2021), 23(8), 933

Achieving real-time inference is one of the major issues in contemporary neural network applications, as complex algorithms are frequently being deployed to mobile devices that have constrained storage ... [more ▼]

Achieving real-time inference is one of the major issues in contemporary neural network applications, as complex algorithms are frequently being deployed to mobile devices that have constrained storage and computing power. Moving from a full-precision neural network model to a lower representation by applying quantization techniques is a popular approach to facilitate this issue. Here, we analyze in detail and design a 2-bit uniform quantization model for Laplacian source due to its significance in terms of implementation simplicity, which further leads to a shorter processing time and faster inference. The results show that it is possible to achieve high classification accuracy (more than 96% in the case of MLP and more than 98% in the case of CNN) by implementing the proposed model, which is competitive to the performance of the other quantization solutions with almost optimal precision. [less ▲]

Fractional Linear Prediction Toolbox for MATLAB

in Proc. of 21th International Carpathian Control Conference (ICCC) (2020, October)

This paper presents the Fractional Linear Prediction (FLP) Toolbox implemented in MATLAB with a supporting livescript interface that offers an user-friendly environment for the prediction of one ... [more ▼]

This paper presents the Fractional Linear Prediction (FLP) Toolbox implemented in MATLAB with a supporting livescript interface that offers an user-friendly environment for the prediction of one-dimensional signals. Two versions of the FLP are implemented in the toolbox and presented here. While the first approach is using the “full” memory (the whole history of the signal), the second approach uses the “restricted” memory (two, three or four previous samples). Both FLP approaches are compared to the standard linear prediction and their performance is evaluated in examples using a test signal (sine wave signal), and a real-data signal (speech signal) as the input. [less ▲]

Speech Based Estimation of Parkinson’s Disease Using Gaussian Processes and Automatic Relevance Determination

in Neurocomputing (2020), 401

Parkinson’s disease is a progressive neurodegenerative disorder often accompanied by impairment in articulation, phonation, prosody and fluency of speech. In fact, speech impairment is one of the earliest ... [more ▼]

Parkinson’s disease is a progressive neurodegenerative disorder often accompanied by impairment in articulation, phonation, prosody and fluency of speech. In fact, speech impairment is one of the earliest Parkinson’s disease symptoms, and may be used for early diagnosis. We present an experimental study of identification of Parkinson’s disease and assessment of disease progress from speech using Gaussian processes, which is further combined with Automatic Relevance Determination (ARD) for efficient feature selection. Hyperparameters of ARD covariance functions are learned for each individual feature; therefore, can be used for evaluation of their importance. In that way only a small subset of highly relevant acoustic features is selected, leading to models with better performance and lower complexity. The performance of the proposed method was assessed on two datasets: Parkinson’s disease detection dataset, which contains a range of biomedical voice measurements obtained from 31 subjects, 23 of them suffering from Parkinson’s disease and 8 healthy subjects; and Parkinson’s telemonitoring dataset, containing biomedical voice measurements collected from 42 Parkinson’s disease patients for estimation of the disease progress. Gaussian process classification with automatic relevance determination is able to successfully discriminate between Parkinson’s disease patients and healthy controls with 96.92% accuracy, outperforming Support Vector Machines and decision tree ensembles (random forests, boosted and bagged decision trees). The usability of Gaussian processes is further confirmed in regression task for tracking the progress of the disease. [less ▲]

Audio signal processing using fractional linear prediction

in Mathematics (2019), 7(7),

Fractional linear prediction (FLP), as a generalization of conventional linear prediction (LP), was recently successfully applied in different fields of research and engineering, such as biomedical signal ... [more ▼]

Fractional linear prediction (FLP), as a generalization of conventional linear prediction (LP), was recently successfully applied in different fields of research and engineering, such as biomedical signal processing, speech modeling and image processing. The FLP model has a similar design as the conventional LP model, i.e., it uses a linear combination of “fractional terms” with different orders of fractional derivative. Assuming only one “fractional term” and using limited number of previous samples for prediction, FLP model with “restricted memory” is presented in this paper and the closed-form expressions for calculation of FLP coefficients are derived. This FLP model is fully comparable with the widely used low-order LP, as it uses the same number of previous samples, but less predictor coefficients, making it more efficient. Two different datasets, MIDI Aligned Piano Sounds (MAPS) and Orchset, were used for the experiments. Triads representing the chords composed of three randomly chosen notes and usual Western musical chords (both of them from MAPS dataset) served as the test signals, while the piano recordings from MAPS dataset and orchestra recordings from the Orchset dataset served as the musical signal. The results show enhancement of FLP over LP in terms of model complexity, whereas the performance is comparable. [less ▲]

Optimal fractional linear prediction with restricted memory

in IEEE Signal Processing Letters (2019), 26(5), 760-764

Linear prediction is extensively used in modeling, compression, coding, and generation of speech signal. Various formulations of linear prediction are available, both in time and frequency domain, which ... [more ▼]

Linear prediction is extensively used in modeling, compression, coding, and generation of speech signal. Various formulations of linear prediction are available, both in time and frequency domain, which start from different assumptions but result in the same solution. In this letter, we propose a novel, generalized formulation of the optimal low-order linear prediction using the fractional (non-integer) derivatives. The proposed fractional derivative formulation allows for the definition of predictor with versatile behavior based on the order of fractional derivative. We derive the closed-form expressions of the optimal fractional linear predictor with restricted memory, and prove that the optimal first-order and the optimal second-order linear predictors are only its special cases. Furthermore, we empirically prove that the optimal order of fractional derivative can be approximated by the inverse of the predictor memory, and thus, it is a priori known. Therefore, the complexity is reduced by optimizing and transferring only one predictor coefficient, i.e., one parameter less in comparison to the second-order linear predictor, at the same level of performance. [less ▲]

An efficient two-digit adaptive delta modulation for Laplacian source coding

in International Journal of Electronics (2019), 106(7), 1085-1100

Delta Modulation (DM) is a simple waveform coding algorithm used mostly when timely data delivery is more important than the transmitted data quality. While the implementation of DM is fairly simple and ... [more ▼]

Delta Modulation (DM) is a simple waveform coding algorithm used mostly when timely data delivery is more important than the transmitted data quality. While the implementation of DM is fairly simple and inexpensive, it suffers from several limitations, such as slope overload and granular noise, which can be overcome using Adaptive Delta Modulation (ADM). This paper presents novel 2-digit ADM with six-level quantization using variable-length coding, for encoding the time-varying signals modelled by Laplacian distribution. Two variants of quantizer are employed, distortion-constrained quantizer which is optimally designed for minimal mean-squared error (MSE), and rate-constrained quantizer, which is suboptimal in the minimal MSE sense, but enables minimal loss in SQNR for the target bit rate. Experimental results using real speech signal are provided, indicating that the proposed configuration outperforms the baseline ADM algorithms, including Constant Factor Delta Modulation (CFDM), Continuously Variable Slope Delta Modulation (CVSDM), 2-digit and 2-bit ADM, and operates in a much wider dynamic range. [less ▲]

Forward Adaptive Laplacian Source Coding Based on Restricted Quantization

in Information Technology and Control (2018), 47(2), 209-219

A novel solution for Laplacian source coding based on three-level quantization is proposed in this paper. The restricted three-level quantizer is designed by assuming the restricted Laplacian distribution ... [more ▼]

A novel solution for Laplacian source coding based on three-level quantization is proposed in this paper. The restricted three-level quantizer is designed by assuming the restricted Laplacian distribution of the input signal. Quantizer and Huffman encoder are jointly designed. Forward adaptive scheme was employed, where the adaptation to the signal variance (power) was performed on frame-by frame basis. We employ switched model that consists of two restricted quantizers having unequal support regions. The simulation results (measured as SQNR) of the proposed scheme with a switched restricted three-level quantizer are compared to the cases when it involves three-level unrestricted quantizer and the Lloyd-Max quantizers having N=2 and N=4 levels. It is shown that the proposed solution offers performance comparable to the one of N=4 levels Lloyd-Max’s baseline with large savings in bit rate, while outperforming two other baselines. [less ▲]

Dual-mode quasi-logarithmic quantizer with embedded G.711 codec

in Journal of Electrical Engineering (2018), 69(1), 46-51

The G.711 codec has been accepted as a standard for high quality coding in many applications. A dual-mode quantizer, which combines the nonlinear logarithmic quantizer for restricted input signals and G ... [more ▼]

The G.711 codec has been accepted as a standard for high quality coding in many applications. A dual-mode quantizer, which combines the nonlinear logarithmic quantizer for restricted input signals and G.711 quantizer for unrestricted input signals is proposed in this paper. The parameters of the proposed quantizer are optimized, where the minimal distortion is used as the criterion. It is shown that the optimized version of the proposed quantizer provides 5.4 dB higher SQNR (Signal to Quantization Noise Ratio) compared to G.711 quantizer, or equivalently it performs savings in the bit rate of approximately 0.9 bit/sample for the same signal quality. Although the complexity is slightly increased, we believe that due to the superior performance it can be successfully implemented for high-quality quantization. [less ▲]

Sentiment Analysis of Microblogs Using Multilayer Feed-forward Artificial Neural Networks

in Computing and Informatics (2017), 36(5), 11271142

Sentiment analysis aims to extract public opinion on a particular topic and microblogs, especially Twitter as the most influential platform, represent a significant source of information. The application ... [more ▼]

Sentiment analysis aims to extract public opinion on a particular topic and microblogs, especially Twitter as the most influential platform, represent a significant source of information. The application to microblogs has to cope with difficulties, such as informal language with abbreviations, internet jargons, emoticons, hashtags that do not appear in conventional text documents. Sentiment analysis technique for microblogs based on a feed-forward artificial neural network (ANN) with sigmoid activation function is proposed in this paper and compared to machine learning approaches, i.e. Multinomial Naive Bayes, Support Vector Machines and Maximum Entropy. Experiments were performed on Stanford Twitter Sentiment corpus, a balanced dataset which contains noisy training labels weakly annotated using emoticons as sentiment indicators; and SemEval-2014 Task 9 corpus, an unbalanced dataset which contains manually annotated training examples. The obtained results show that ANN produces superior or at least comparable results to state-of-the-art machine learning techniques. [less ▲]