EEG-Based BCIs on Motor Imagery Paradigm Using Wearable Technologies: A Systematic Review.

In recent decades, the automatic recognition and interpretation of brain waves acquired by electroencephalographic (EEG) technologies have undergone remarkable growth, leading to a consequent rapid development of brain-computer interfaces (BCIs). EEG-based BCIs are non-invasive systems that allow communication between a human being and an external device interpreting brain activity directly. Thanks to the advances in neurotechnologies, and especially in the field of wearable devices, BCIs are now also employed outside medical and clinical applications. Within this context, this paper proposes a systematic review of EEG-based BCIs, focusing on one of the most promising paradigms based on motor imagery (MI) and limiting the analysis to applications that adopt wearable devices. This review aims to evaluate the maturity levels of these systems, both from the technological and computational points of view. The selection of papers has been performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), leading to 84 publications considered in the last ten years (from 2012 to 2022). Besides technological and computational aspects, this review also aims to systematically list experimental paradigms and available datasets in order to identify benchmarks and guidelines for the development of new applications and computational models.

Behavior and Task Classification Using Wearable Sensor Data: A Study across Different Ages.

In this paper, we face the problem of task classification starting from physiological signals acquired using wearable sensors with experiments in a controlled environment, designed to consider two different age populations: young adults and older adults. Two different scenarios are considered. In the first one, subjects are involved in different cognitive load tasks, while in the second one, space varying conditions are considered, and subjects interact with the environment, changing the walking conditions and avoiding collision with obstacles. Here, we demonstrate that it is possible not only to define classifiers that rely on physiological signals to predict tasks that imply different cognitive loads, but it is also possible to classify both the population group age and the performed task. The whole workflow of data collection and analysis, starting from the experimental protocol, data acquisition, signal denoising, normalization with respect to subject variability, feature extraction and classification is described here. The dataset collected with the experiments together with the codes to extract the features of the physiological signals are made available for the research community.

Comparing online cognitive load on mobile versus PC-based devices.

Navigating the web represents a complex cognitive activity that requires effective integration of different stimuli and the correct functioning of numerous cognitive abilities (including attention, perception, and working memory). Despite the potential relevance of the topic, numerous limitations are present throughout the literature about the cognitive load during online activities. The main aim of this study is to investigate cognitive load during comprehension and information-seeking tasks. In particular, we here focus on the comparison of the cognitive load required while performing those tasks using mobile or PC-based devices. This topic has become even more crucial due to the massive adoption of smart working and distance learning during the COVID-19 pandemic. A great effort is nowadays devoted to the detection and quantification of stressful states induced by working and learning activities. Continuous stress and excessive cognitive load are two of the main causes of mental and physical illnesses such as depression or anxiety. Cognitive load was measured through electroencephalography (EEG), acquired via a low-cost wireless EEG headset. Two different tasks were considered: reading comprehension (CO) of online text and online information-seeking (IS). Moreover, two experimental conditions were compared, administering the two tasks using mobile (MB) and desktop (PC) devices. Eleven participants were involved in each experimental condition, MB and PC, performing both the tasks on the same device, for a total of twenty-two people, recruited from students, researchers, and employees of the university. The following two research questions were investigated: Q1: Is there a difference in the cognitive load while performing the comprehension and the information-seeking tasks? Q2: Does the adopted device influence the cognitive load? The results obtained show that the baseline (BL) requires the lower cognitive load in both the conditions, while in IS task, the requirement reaches its highest value, especially using a mobile phone. In general, the power of all the brain wave bands increased in all conditions (MB and PC) during the two tasks (CO and IS), except for alpha, which is usually high in a state of relaxation and low cognitive load. People include website navigation into their daily routines, and for this, it is important to create an interaction that is as easy and barrier-free as possible. An effective design allows a user to focus on interesting information: many website architectures, instead, are an obstacle to be overcome; they impose a high cognitive load and poor user experience. All these aspects draw cognitive resources away from the user's primary task of finding and comprehending the site's information. Having information about how the cognitive load varies based on the device adopted and the considered task can provide useful indicators in this direction. This work suggests that using an EEG low-cost wearable device could be useful to quantify the cognitive load induced, allowing the development of new experiments to analyse these dependencies deeper, and to provide suggestions for better interaction with the web.

Benchmark dataset of memes with text transcriptions for automatic detection of multi-modal misogynistic content.

In this paper we present a benchmark dataset generated as part of a project for automatic identification of misogyny within online content, which focuses in particular on memes. The benchmark here described is composed of 800 memes collected from the most popular social media platforms, such as Facebook, Twitter, Instagram and Reddit, and consulting websites dedicated to collection and creation of memes. To gather misogynistic memes, specific keywords that refer to misogynistic content have been considered as search criterion, considering different manifestations of hatred against women, such as body shaming, stereotyping, objectification and violence. In parallel, memes with no misogynist content have been manually downloaded from the same web sources. Among all the collected memes, three domain experts have selected a dataset of 800 memes equally balanced between misogynistic and non-misogynistic ones. This dataset has been validated through a crowdsourcing platform, involving 60 subjects for the labelling process, in order to collect three evaluations for each instance. Two further binary labels have been collected from both the experts and the crowdsourcing platform, for memes evaluated as misogynistic, concerning aggressiveness and irony. Finally for each meme, the text has been manually transcribed. The dataset provided is thus composed of the 800 memes, the labels given by the experts and those obtained by the crowdsourcing validation, and the transcribed texts. This data can be used to approach the problem of automatic detection of misogynistic content on the Web relying on both textual and visual cues, facing phenomenons that are growing every day such as cybersexism and technology-facilitated violence.

Health Literacy, Vaccine Confidence and Influenza Vaccination Uptake among Nursing Home Staff: A Cross-Sectional Study Conducted in Tuscany.

The aim of this cross-sectional study is to address whether health literacy (HL) and vaccine confidence are related with influenza vaccination uptake among staff of nursing homes (NHs). It was conducted in Tuscany (Italy) in autumn 2018, including the staff of 28 NHs. A questionnaire was used to collect individual data regarding influenza vaccination in 2016-2017 and 2017-2018 seasons; the intention to be vaccinated in 2018-2019; as well as demographic, educational, and health information. It included also the Italian Medical Term Recognition (IMETER) test to measure HL and eight Likert-type statements to calculate a Vaccine Confidence Index (VCI). The number of employees that fulfilled the questionnaire was 710. The percentage of influenza vaccination uptake was low: only 9.6% got vaccinated in 2016-2017 and 2017-2018 and intended to vaccinate in 2018-2019. The VCI score and the IMETER-adjusted scores were weakly correlated (Rho = 0.156). At the multinomial logistic regression analysis, the VCI was a positive predictor of vaccination uptake. In conclusion, vaccine confidence is the strongest predictor of influenza vaccination uptake among the staff of NHs. The development of an adequate vaccine literacy measurement tool could be useful to understand whether skills could be related to vaccine confidence.

Clustering the Brain With "CluB": A New Toolbox for Quantitative Meta-Analysis of Neuroimaging Data.

In this paper we describe and validate a new coordinate-based method for meta-analysis of neuroimaging data based on an optimized hierarchical clustering algorithm: CluB (Clustering the Brain). The CluB toolbox permits both to extract a set of spatially coherent clusters of activations from a database of stereotactic coordinates, and to explore each single cluster of activation for its composition according to the cognitive dimensions of interest. This last step, called "cluster composition analysis," permits to explore neurocognitive effects by adopting a factorial-design logic and by testing the working hypotheses using either asymptotic tests, or exact tests either in a classic inference, or in a Bayesian-like context. To perform our validation study, we selected the fMRI data from 24 normal controls involved in a reading task. We run a standard random-effects second level group analysis to obtain a "Gold Standard" of reference. In a second step, the subject-specific reading effects (i.e., the linear t-contrast "reading > baseline") were extracted to obtain a coordinates-based database that was used to run a meta-analysis using both CluB and the popular Activation Likelihood Estimation method implemented in the software GingerALE. The results of the two meta-analyses were compared against the "Gold Standard" to compute performance measures, i.e., sensitivity, specificity, and accuracy. The GingerALE method obtained a high level of accuracy (0.967) associated with a high sensitivity (0.728) and specificity (0.971). The CluB method obtained a similar level of accuracy (0.956) and specificity (0.969), notwithstanding a lower level of sensitivity (0.14) due to the lack of prior Gaussian transformation of the data. Finally, the two methods obtained a good-level of concordance (AC1 = 0.93). These results suggested that methods based on hierarchical clustering (and post-hoc statistics) and methods requiring prior Gaussian transformation of the data can be used as complementary tools, with the GingerALE method being optimal for neurofunctional mapping of pooled data according to simpler designs, and the CluB method being preferable to test more specific, and localized, neurocognitive hypotheses according to factorial designs.

Computational Methods for Resting-State EEG of Patients With Disorders of Consciousness.

Patients who survive brain injuries may develop Disorders of Consciousness (DOC) such as Coma, Vegetative State (VS) or Minimally Conscious State (MCS). Unfortunately, the rate of misdiagnosis between VS and MCS due to clinical judgment is high. Therefore, diagnostic decision support systems aiming to correct any differentiation between VS and MCS are essential for the characterization of an adequate treatment and an effective prognosis. In recent decades, there has been a growing interest in the new EEG computational techniques. We have reviewed how resting-state EEG is computationally analyzed to support differential diagnosis between VS and MCS in view of applicability of these methods in clinical practice. The studies available so far have used different techniques and analyses; it is therefore hard to draw general conclusions. Studies using a discriminant analysis with a combination of various factors and reporting a cut-off are among the most interesting ones for a future clinical application.

Databases for evaluating interferences between affective content and image quality.

The two databases here described were generated to evaluate the role of affective content while assessing image quality (Corchs et al., 2018) [1]. The databases are composed of images JPEG-compressed together with the subjective quality scores collected during psychophysical experiments. To reduce interferences in quality perception due to image semantic, we have restricted the semantic content, choosing only close-ups of face images, and we have considered only two emotion categories (happy and sad). We have selected 23 images with happy faces and 23 images with sad faces of high quality. For what concerns image quality we have considered JPEG-distortion with 4 levels of compression, corresponding to q-factors 10, 15, 20, 30. The first image database, hereafter called MMSP-FaceA, is thus composed of 230 images (23+23) × 5 quality levels (including the original high quality pristine images). To better consider only interferences in quality perception due to affective content, we have generated a second image database where the background of images belonging to MMSP-FaceA has been cut off. This second image database is labelled as MMSP-FaceB. Psychophysical experiments were conducted, on a controlled web-based interface, where participants rated the image quality of the two databases in a five point scale. The two final databases MMSP-FaceA and MMSP-FaceB are thus composed of 230 images each, together with the raw quality scores assigned by the observers, and are available at our laboratory web site: www.mmsp.unimib.it/download.

Corrigendum: Computational Methods for Resting-State EEG of Patients With Disorders of Consciousness.

[This corrects the article DOI: 10.3389/fnins.2019.00807.].

The impact of overdiagnosis on thyroid cancer epidemic in Italy,1998-2012.

AIMS: In Italy, incidence rates of thyroid cancer (TC) are among the highest worldwide with substantial intracountry heterogeneity. The aim of the study was to examine time trends of TC incidence in Italy and to estimate the proportion of TC cases potentially attributable to overdiagnosis. METHODS: Data on TC cases reported to Italian cancer registries during 1998-2012 aged <85 years were included. Age-standardised incidence rates (ASR) were computed by sex, period, and histology. TC overdiagnosis was estimated by sex, period, age, and Italian region. RESULTS: In Italy between 1998-2002 and 2008-2012, TC ASR increased of 74% in women (from 16.2 to 28.2/100,000) and of 90% in men (from 5.3 to 10.1/100,000). ASR increases were nearly exclusively due to papillary TC (+91% in women, +120% in men). In both sexes, more than three-fold differences emerged between regions with highest and lowest ASR. Among TC cases diagnosed in 1998-2012 in Italy, we estimated that overdiagnosis accounted for 75% of cases in women and 63% in men and increased over the study period leading to overdiagnosis of 79% in women and 67% in men in 2008-2012. Notably, overdiagnosis was over 80% among women aged <55 years, and substantial variations were documented across Italian regions, in both genders. CONCLUSION(S): Incidence rates of TC are steadily increasing in Italy and largely due to overdiagnosis. These findings call for an update of thyroid gland examination practices in the asymptomatic general population, at national and regional levels.

Predicting Complexity Perception of Real World Images.

The aim of this work is to predict the complexity perception of real world images. We propose a new complexity measure where different image features, based on spatial, frequency and color properties are linearly combined. In order to find the optimal set of weighting coefficients we have applied a Particle Swarm Optimization. The optimal linear combination is the one that best fits the subjective data obtained in an experiment where observers evaluate the complexity of real world scenes on a web-based interface. To test the proposed complexity measure we have performed a second experiment on a different database of real world scenes, where the linear combination previously obtained is correlated with the new subjective data. Our complexity measure outperforms not only each single visual feature but also two visual clutter measures frequently used in the literature to predict image complexity. To analyze the usefulness of our proposal, we have also considered two different sets of stimuli composed of real texture images. Tuning the parameters of our measure for this kind of stimuli, we have obtained a linear combination that still outperforms the single measures. In conclusion our measure, properly tuned, can predict complexity perception of different kind of images.

Molecular characterization of HIV-1 Nef and ACOT8 interaction: insights from in silico structural predictions and in vitro functional assays.

HIV-1 Nef interacts with several cellular proteins, among which the human peroxisomal thioesterase 8 (ACOT8). This interaction may be involved in the endocytosis regulation of membrane proteins and might modulate lipid composition in membrane rafts. Nef regions involved in the interaction have been experimentally characterized, whereas structural details of the ACOT8 protein are unknown. The lack of structural information hampers the comprehension of the functional consequences of the complex formation during HIV-1 infection. We modelled, through in silico predictions, the ACOT8 structure and we observed a high charge complementarity between Nef and ACOT8 surfaces, which allowed the identification of the ACOT8 putative contact points involved in the interaction. The predictions were validated by in vitro assays through the development of ACOT8 deletion mutants. Coimmunoprecipitation and immunofluorescence analyses showed that ACOT8 Arg(45)-Phe(55) and Arg(86)-Pro(93) regions are involved in Nef association. In addition, K91S mutation abrogated the interaction with Nef, indicating that Lys(91) plays a key role in the interaction. Finally, when associated with ACOT8, Nef may be preserved from degradation. These findings improve the comprehension of the association between HIV-1 Nef and ACOT8, helping elucidating the biological effect of their interaction.

Adaptive Skin Classification Using Face and Body Detection.

In this paper, we propose a skin classification method exploiting faces and bodies automatically detected in the image, to adaptively initialize individual ad hoc skin classifiers. Each classifier is initialized by a face and body couple or by a single face, if no reliable body is detected. Thus, the proposed method builds an ad hoc skin classifier for each person in the image, resulting in a classifier less dependent from changes in skin color due to tan levels, races, genders, and illumination conditions. The experimental results on a heterogeneous data set of labeled images show that our proposal outperforms the state-of-the-art methods, and that this improvement is statistically significant.

Coding polymorphisms in the parkin gene and susceptibility to Parkinson disease.

BACKGROUND: Mutations in the parkin gene, an E3 protein-ubiquitin ligase, cause autosomal recessive early-onset Parkinson disease (PD). The role of polymorphisms in the parkin gene as risk factors for PD is still unclear, as the results in the literature are contradictory. PATIENTS: We compared the allele and genotype frequencies of the Ser167Asn, Arg366Trp, Val380Leu, and Asp394Asn polymorphisms in 194 patients with PD (92 familial and 102 sporadic) and 125 control subjects. RESULTS: Homozygous Val380 was significantly associated with sporadic PD (P =.008). There was also a trend toward an association of homozygous Asp394 with familial PD (P =.07). CONCLUSIONS: Some parkin polymorphisms appear to be risk factors for sporadic or familial PD. The functional effects of these coding polymorphisms need to be established, and further studies on parkin polymorphisms in PD should be undertaken.

How much phenotypic variation can be attributed to parkin genotype?

To establish phenotype-genotype correlations in early-onset parkinsonism, we have compared the phenotype of a large series of 146 patients with and 250 patients without parkin mutations. Although no single sign distinguished the groups, patients with mutations had significantly earlier and more symmetrical onset, dystonia more often at onset and hyperreflexia, slower progression of the disease, and a tendency toward a greater response to levodopa despite lower doses. After forward stepwise multiple logistic regression analysis, dystonia at onset and brisk reflexes were not longer significantly different but were correlated with age at onset rather than the presence of the parkin mutation. Age at onset in carriers of parkin mutations varied as did the rate of progression of the disease: the younger the age at onset the slower the evolution. The genotype influenced the phenotype: carriers of at least one missense mutation had a higher United Parkinson's Disease Rating Scale motor score than those carrying two truncating mutations. The localization of the mutations was also important because missense mutations in functional domains of parkin resulted in earlier onset. Patients with a single heterozygous mutation had significantly later and more asymmetrical onset and more frequent levodopa-induced fluctuations and dystonia than patients with two mutations.

Association study between iron-related genes polymorphisms and Parkinson's disease.

We have conducted a case-control study in order to test for an association between 8 intragenic polymorphisms of 5 iron-related genes (transferrin, transferrin receptor1, HFE, frataxin and lactoferrin) and Parkinson disease. Comparison of genotypes and allele frequencies did not differ significantly between cases and controls for all studied polymorphisms except the G258S transferrin polymorphism, for which a higher frequency of the G allele was found among cases (p=0.033), particularly among cases with onset older than 60 (p=0.0017) and with negative family history (p=0.022). This finding suggests that genetic variations in the control of iron metabolism may contribute to the pathogenesis of the disease.