This study investigates the integration of multiple biological signals to assess the impact of sleep deprivation on attention levels. Electrocardiogram (ECG), electroencephalogram (EEG), and electrooculogram (EOG) data from sleep-deprived patients were analyzed with performance outcomes from the Psychomotor Vigilance Test (PVT), which measures response times. The primary objective was to develop a robust predictive model for the level of drowsiness based on these signals. By leveraging machine learning models, the study demonstrated the feasibility of signal-based assessments for predicting drowsiness levels. Random Forest achieved the highest accuracy when using reaction times as the true labels. It also showed promising agreement with the subjective evaluation of the alertness levels, highlighting conditions where the individuals may risk to underestimate their drowsiness. The results underscore the potential of biological signals to improve understanding of sleep deprivation’s impact on cognitive performance and potentially contribute to develop robust drowsiness detection systems for practical contexts.

In-ear EEG has recently emerged as a promising avenue to assess cognitive workload using minimally obtrusive sensors, thus promoting continuous and ubiquitous health monitoring. However, concerns around the quality and representativeness of data collected with this new technology need further investigations. In this work, we utilize a dataset related to a participant engaged in various mathematical tasks while wearing an in-ear EEG device. We apply signal processing techniques and feature extraction methodologies to analyze the EEG data. Feature vectors were constructed from each data segment, and subsequently used to train various machine learning classifiers to discriminate between different levels of cognitive workload. Moreover, we investigate the effectiveness of feature selection methods, to reduce the dimensionality of the feature space and potentially improve classifier performance. The results indicate that in-ear EEG, together with proper processing in terms of feature selection and machine learning, can adequately differentiate cognitive workload levels. Our findings proved the convenience of carrying on the investigation of this new kind of technology to promote a healthcare service closer to patients.

Smart agriculture has seen impressive progresses in monitoring the quality of the crop and early detecting the onset of pathogens.However, this is typically achieved through smart, expensive, and energy-demanding robots and autonomous systems.We propose an AI-empowered portable low-cost shortwave near-infrared spectroscopy (sw-NIRS) solution that allows non-destructive measurements from plants and vegetables.In this pilot study, we specifically targeted an orange fruit and showed that it is possible to classify its different parts through sw-NIRS in the range 1350-2150 nm by using AI models, exceeding 97% accuracy.Also, we explored the minimum amount of energy needed to reach such high classification performance.In the future, we aim to extend this investigation to other targets (e.g., bean plants), to develop AI architectures to more accurately model the physiological conditions of the target, and to create a network of sw-NIRS sensors to simultaneously monitor a largescale crop.

The classification of EEG during motor imagery (MI) represents a challenging task in neuro-rehabilitation. In 2016, a deep learning (DL) model called EEGNet (based on CNN) and its variants attracted much attention for their ability to reach 80% accuracy in a 4-class MI classification. However, they can poorly explain their output decisions, preventing them from definitely solving questions related to inter-subject variability, generalization, and optimal classification. In this paper, we propose vEEGNet, a new model based on EEGNet, whose objective is now two-fold: it is used to classify MI, but also to reconstruct (and eventually generate) EEG signals. The work is still preliminary, but we are able to show that vEEGNet is able to classify 4 types of MI with performances at the state of the art, and, more interestingly, we found out that the reconstructed signals are consistent with the so-called motor-related cortical potentials, very specific and well-known motorrelated EEG patterns. Thus, jointly training vEEGNet to both classify and reconstruct EEG might lead it, in the future, to decrease the inter-subject performance variability, and also to generate new EEG samples to augment small datasets to improve classification, with a consequent strong impact on neuro-rehabilitation.

This paper investigates data correlation in remote sensing networks and how it can be characterized through diverse models quantifying the Value of Information (VoI), a metric that describes how informative the data transmitted by the sensors are. For each sensor, the VoI evaluations comprise the average node-specific Age of Information (AoI), the average cost spent for sending updates, and the AoI of neighbor nodes, assumed to be correlated sources of information and therefore benefiting the VoI of other sensors nearby. We discuss how this metric can be tracked through a two-dimensional Markov chain, but we also show how this representation can be simplified by including the impact of neighbor nodes within the transition probabilities, so as to obtain a simpler model that gives the same insight in terms of VoI evaluations.

This paper investigates the decision making process aided by machine learning for biomedical problems and how to improve it through meta assessments of the most relevant features. Classification algorithms are usually trained and exploited with high dimensional datasets (i.e., with an extremely large number of features), which is inefficient and costly. It would be beneficial to identify the most meaningful features that contribute the most to assigning a category to a subject, and in particular, diagnosing a pathological condition. A helpful support can come from cooperative game theory, through the computation of the Shapley value, an indicator of desirable properties according to which the players, in our case the input features, can be ranked. We apply such a framework to a supervised machine learning scenario of a random forest tree classifier applied to heart disease detection. From a publicly available dataset, we identify the most relevant features that can affect the decision, thus obtaining practical guidelines for a compact yet efficient description based on an analytical rationale.

The problem of multi-agent robotic survey of an unknown area is approached through a game theoretic frame-work. This is meant to enable cooperation in the group of robotic explorers reflecting their common objectives to minimize the effort in the surveying task, without requiring expensive exchanges of signaling. The game theoretic approach can be applied to avoid any preliminary planning, but just exploiting the ability of the robots to take smart actions based on the state of the environment and the behaviors of other neighboring agents. We discuss how the choice of different utility functions can improve the collaboration among the robots and lead to more efficient results.

The accurate detection of motor imagery (MI) from electroencephalography (EEG) is a fundamental, as well as challenging, task to provide reliable control of robotic devices to support people suffering from neuro-motor impairments, e.g., in brain-computer interface (BCI) applications. Recently, deep learning approaches have been able to extract subject-independent features from EEG, to cope with its poor SNR and high intra-subject and cross-subject variability. In this paper, we first present a review of the most recent studies using deep learning for MI classification, with particular attention to their cross-subject performance. Second, we propose DynamicNet, a Python-based tool for quick and flexible implementations of deep learning models based on convolutional neural networks. We showcase the potentiality of DynamicNet by implementing EEGNet, a well-established architecture for effective EEG classification. Finally, we compare its performance with the filter bank common spatial pattern (FBCSP) in a 4-class MI task (data from a public dataset). To infer cross-subject classification performance, we applied three different cross-validation schemes. From our results, we show that EEGNet implemented with DynamicNet outperforms FBCSP by about 25 %, with a statistically significant difference when cross-subject validation schemes are applied. We conclude that deep learning approaches might be particularly helpful to provide higher cross-subject classification performance in multiclass MI classification scenarios. In the future, it is expected to improve DynamicNet to implement new architectures to further investigate cross-subject classification of MI tasks in real-world scenarios.

Wireless body area networks (WBANs) are becoming increasingly popular as they allow individuals to continuously monitor their vitals and physiological parameters remotely from the hospital. With the spread of the SARS-CoV-2 pandemic, the availability of portable pulse-oximeters and wearable heart rate detectors has boomed in the market. At the same time, in 2020 we assisted to an unprecedented increase of healthcare breaches, revealing the extreme vulnerability of the current generation of WBANs. Therefore, the development of new security protocols to ensure data protection, authentication, integrity and privacy within WBANs are highly needed. Here, we targeted a WBAN collecting ECG signals from different sensor nodes on the individual's body, we extracted the inter-pulse interval (i.e., R-R interval) sequence from each of them, and we developed a new information theoretic key agreement protocol that exploits the inherent randomness of ECG to ensure authentication between sensor pairs within the WBAN. After proper pre-processing, we provide an analytical solution that ensures robust authentication; we provide a unique information reconciliation matrix, which gives good performance for all ECG sensor pairs; and we can show that a relationship between information reconciliation and privacy amplification matrices can be found. Finally, we show the trade-off between the level of security, in terms of key generation rate, and the complexity of the error correction scheme implemented in the system.

Recent evidence has revealed cross-frequency coupling and, particularly, phase-amplitude coupling (PAC) as an important strategy for the brain to accomplish a variety of high-level cognitive and sensory functions. However, decoding PAC is still challenging. This contribution presents REPAC, a reliable and robust algorithm for modeling and detecting PAC events in EEG signals. First, we explain the synthesis of PAC-like EEG signals, with special attention to the most critical parameters that characterize PAC, i.e., SNR, modulation index, duration of coupling. Second, REPAC is introduced in detail. We use computer simulations to generate a set of random PAC-like EEG signals and test the performance of REPAC with regard to a baseline method. REPAC is shown to outperform the baseline method even with realistic values of SNR, e.g., −10 dB. They both reach accuracy levels around 99%, but REPAC leads to a significant improvement of sensitivity, from 20.11% to 65.21%, with comparable specificity (around 99%). REPAC is also applied to a real EEG signal showing preliminary encouraging results.

Internet of Things is rapidly spreading across several fields, including healthcare, posing relevant questions related to communication capabilities, energy efficiency and sensors unobtrusiveness. Particularly, in the context of recognition of gestures, e.g., grasping of different objects, brain and muscular activity could be simultaneously recorded via EEG and EMG, respectively, and analyzed to identify the gesture that is being accomplished, and the quality of its performance. This paper proposes a new algorithm that aims (i) to robustly extract the most relevant features to classify different grasping tasks, and (ii) to retain the natural meaning of the selected features. This, in turn, gives the opportunity to simplify the recording setup to minimize the data traffic over the communication network, including Internet, and provide physiologically significant features for medical interpretation. The algorithm robustness is ensured both by consensus clustering as a feature selection strategy, and by nested cross-validation scheme to evaluate its classification performance. Although Feature Selection with Consensus (FeSC) implements a very robust architecture for feature selection and classification, results are still negatively affected by the limited size of the dataset. In the future, further investigations could determine to what extent size could cause a drop in the performance of FeSC in this and other gesture recognition applications.

In this paper, we propose a new method for jointly compressing EEG and EMG biosignals based on the so-called cortico-muscular coherence, a function that takes into account the simultaneous frequency changes of the brain and the muscles activity, and can be used, e.g., to classify different kinds of movement. It is shown that this method increases the achievable compression rate compared to transmitting EEG and EMG samples separately, while trading-off with the accuracy of the classification. This can be exploited in several kinds of life and health applications e.g., motor rehabilitation and drivers attention monitoring; it could be especially useful for low-power wireless technologies, such as Bluetooth Low Energy or IEEE 802.15.6, whose transmission resources are limited.

This work presents an innovative application of the well-known concept of cortico-muscular coherence for the classification of various motor tasks, i.e., grasps of different kinds of objects. Our approach can classify objects with different weights (motor-related features) and different surface frictions (haptics-related features) with high accuracy (over 0.8). The outcomes presented here provide information about the synchronization existing between the brain and the muscles during specific activities; thus, this may represent a new effective way to perform activity recognition.

Due to their distributed management, smart grids can be vulnerable to malicious attacks that undermine their cyber security. An adversary can take control of few nodes in the network and spread digital attacks like an infection, whose diffusion is facilitated by the lack of centralized supervision within the smart grid. In this paper, we propose to investigate these phenomena by means of epidemic models applied to cellular automata. We show that the common key parameters of epidemic models, such as the basic reproductive ratio, are also useful in this context to understand the extent of the grid portion that can be compromised. At the same time, the lack of mobility of individuals limits the spreading of the infection. In particular, we evaluate the role of the grid connectivity degree in both containing the epidemics and avoiding its spreading on the entire network, and also increasing the number of nodes that do not get any contact with the cyber attacks.

The paper deals with a specific kind of BCI application implemented with the aim of recovering the reaching ability of mild impaired stroke survivors. The overall idea is to take advantage of the plasticity of the brain to make the subject artificially learn alternative neural paths to control the arm movement again, by-passing the injured area thanks to a BCI system with an EEG-related force provided as a real-time feedback during the training period. Preliminary results have shown improvements in the kinematics of the upper limb motion of a first patient that performed this experimental rehabilitative program. Then, this BCI application is expected to enter soon the daily clinical practise as a useful tool besides the standard rehabilitation therapy.

Motor rehabilitation after stroke injury is highly important since the number of people suffering this disease is constantly increasing. Brain-Computer Interfaces (BCIs) have been recently used in the recovery of motor functions: in particular, the closed loop involving sensorimotor brain rhythms, assist-ive-robot training and proprioceptive feedback in an operant learning fashion might be potentially one of the most effective ways to promote the neural plasticity of the ipsilesional brain hemisphere and to restore motor abilities. This study aimed at implementing such a scheme: one chronic stroke patient was recruited and underwent the experiment using both the damaged and the healthy arm, considered as control during the following analysis. Kinematic and neurophysiological outcomes confirmed the efficacy of this treatment and supported the hypothesis that a contingent force feedback can improve motor functions of the upper limb.