Parkinson’s disease is a progressive neurodegenerative disorder caused by dopaminergic neuron degeneration. Parkinsonian speech impairment is one of the earliest presentations of the disease and, along with tremor, is suitable for pre-diagnosis. It is defined by hypokinetic dysarthria and accounts for respiratory, phonatory, articulatory, and prosodic manifestations. The topic of this article targets artificial-intelligence-based identification of Parkinson’s disease from continuous speech recorded in a noisy environment. The novelty of this work is twofold. First, the proposed assessment workflow performed speech analysis on samples of continuous speech. Second, we analyzed and quantified Wiener filter applicability for speech denoising in the context of Parkinsonian speech identification. We argue that the Parkinsonian features of loudness, intonation, phonation, prosody, and articulation are contained in the speech, speech energy, and Mel spectrograms. Thus, the proposed workflow follows a feature-based speech assessment to determine the feature variation ranges, followed by speech classification using convolutional neural networks. We report the best classification accuracies of 96% on speech energy, 93% on speech, and 92% on Mel spectrograms. We conclude that the Wiener filter improves both feature-based analysis and convolutional-neural-network-based classification performances.

Walking is the most basic form of human activity for achieving mobility. As an essential function of the human body, the examination of walking is directed towards the assessment of body mechanics in posture and during movement. This work proposes a wearable smart system for the monitoring and objective evaluation of foot biomechanics during gait. The proposed solution assumes the cross-correlation of the plantar pressure with lower-limb muscular activity, throughout the stance phase of walking. Plantar pressure is acquired with an array of resistive pressure sensors deployed onto a shoe insole along the center of gravity progression line. Lower-limb muscular activity is determined from the electromyogram of the tibialis anterior and gastrocnemius lower limb muscles respectively. Under this scenario, physiological gait assumes the interdependency of plantar pressure on the heel area with activation of the tibialis anterior, as well as plantar pressure on the metatarsal arch/toe area with activation of the gastrocnemius. As such, assessment of gait physiology is performed by comparison of a gait map, formulated based on the footprint–lower-limb muscle cross-correlation results, to a reference gait template. A laboratory proof of concept validates the proposed solution in a test scenario which assumes a normal walking and two pathological walking patterns.

Saliva has gained considerable attention as a diagnostics alternative to blood analyses. A wide spectrum of salivary compounds is correlated to blood concentrations of biomarkers, providing informative and discriminative data regarding the state of health. Intra-oral detection and assessment of food and beverage intake can be correlated and provides valuable information to forecast the formation and modification of salivary biomarkers. In this context, the present work proposes a novel intra-oral optical fiber sensor, developed around an optical coupler topology, and exemplified on the detection and assessment of wine intake, which is accounted for example for the formation of N^ε-carboxymethyllysine Advanced Glycation End-products. A laboratory proof of concept validates the proposed solution on four white and four red wine samples. The novel optical sensor geometry shows good spectral properties, accounting for selectivity with respect to grape-based soft drinks. This enables intra-oral detection and objective quality assessment of wine. Moreover, its implementation exploits the advantages of fiber-optics sensing and facilitates integration into a mouthguard, holding considerable potential for real-time biomedical applications to investigate Advanced Glycation End-products in the saliva and their connection with consumption of wine, for the evaluation of risk factors in diet-related diseases.

Recent advances in deep learning have had a significant impact on biomedical applications, driving precise actions in automated diagnostic processes. However, integrating neural networks into medical devices requires meeting strict requirements regarding computing power, energy efficiency, reconfigurability, and latency, essential conditions for real-time inference. Field-Programmable Gate Array (FPGA) architectures provide a high level of flexibility, performance, and parallel execution, thus making them a suitable option for the real-world implementation of machine learning (ML) and deep learning (DL) models in systems dedicated to the analysis of physiological signals. This paper presents a review of intelligent algorithms for electrocardiogram (ECG) signal classification, including Support Vector Machines (SVMs), Artificial Neural Networks (ANNs), Recurrent Neural Networks (RNNs), Long Short-Term Memory Networks (LSTMs), and Convolutional Neural Networks (CNNs), which have been implemented on FPGA platforms. A comparative evaluation of the performances of these hardware-accelerated solutions is provided, focusing on their classification accuracy. At the same time, the FPGA families used are analyzed, along with the reported performances in terms of operating frequency, power consumption, and latency, as well as the optimization strategies applied in the design of deep learning hardware accelerators. The conclusions emphasize the popularity and efficiency of CNN architectures in the context of ECG signal classification. The study aims to offer a current overview and to support specialists in the field of FPGA design and biomedical engineering in the development of accelerators dedicated to physiological signals analysis.