Miguel Ángel Teruel

Background: Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental disorder in children, characterized by inattention, hyperactivity, and impulsivity that significantly disrupt daily functioning. Comprehensive ADHD assessment requires multimethod and multi-informant evaluation, and objective attention measures can complement this process. Serious game–based assessments may further support this process by combining structured task mechanics and dynamic audiovisual guidance in a child-oriented digital environment. Objective: This study aimed to develop and validate Attention Robots, a serious game designed for assessing behavioral attention indicators in children and adolescents. The goal was to examine whether its outputs corresponded with indicators from the widely used D2-R Test of Attention as preliminary evidence of convergent validity. Methods: In this preliminary cross-sectional convergent validation study, 58 participants (34 male and 24 female individuals) aged 6-18 years, recruited through clinical referral and community sources, completed Attention Robots followed by the D2-R test. The original group variable, assigned by the project psychologists, identified 19 participants in the experimental group and 39 participants in the control group. Attention Robots presents timed rows of robot stimuli and asks participants to identify and select predefined target robots. Spearman and Pearson correlations compared working speed, omissions, commissions, concentration, and precision across both instruments. Results: Primary cross-test correlations were positive. Spearman correlations were observed for working speed (ρ=0.905, 95% CI 0.84-0.94; P<.001), omissions (ρ=0.333, 95% CI 0.08-0.54; P=.01), commissions (ρ=0.496, 95% CI 0.27-0.67; P<.001), concentration (ρ=0.926, 95% CI 0.88-0.96; P<.001), and precision (ρ=0.607, 95% CI 0.41-0.75; P<.001). Pearson correlations showed a similar pattern for working speed (r=0.911, 95% CI 0.85-0.95; P<.001), omissions (r=0.278, 95% CI 0.02-0.50; P=.04), commissions (r=0.396, 95% CI 0.15-0.59; P=.002), concentration (r=0.922, 95% CI 0.87-0.95; P<.001), and precision (r=0.574, 95% CI 0.37-0.72; P<.001). These findings support preliminary convergent correspondence between corresponding behavioral indicators, but they do not establish interchangeability, discriminative validity, diagnostic use, or superior engagement or acceptability. Conclusions: Attention Robots provides preliminary evidence of convergent validity with the D2-R for behavioral attention indicators in children and adolescents. The innovative aspect of this study is the translation of D2-R–based attentional demands into a child-oriented serious game environment with robot-based stimuli, dynamic audiovisual guidance, and structured behavioral data capture. Unlike previous digital or serious game approaches, this study empirically compares game-derived indicators with corresponding indicators from an established attention test rather than evaluating only task feasibility, engagement, or intervention effects. This work contributes to the field by providing a preliminary validation of a serious game format that may support the collection of objective attention-related data in children and adolescents. In applied contexts, Attention Robots could be further developed as a complementary tool within multimethod ADHD assessment, although larger studies are required to evaluate subgroup performance, agreement, diagnostic use, and user experience before clinical adoption.

Automatic detection of Alzheimer’s disease (AD) from speech offers a promising non-invasive tool for early screening. This study presents a comprehensive benchmark of deep learning architectures and spectrogram representations for subject-level AD classification using acoustic features alone. We systematically evaluate three spectrogram types (LMS, PCEN, ERB), three neural models (ResNet18Audio, CRNNBiLSTM with attention, ResNeXtAudio), and a broad range of signal and temporal parameters, totaling 3888 experimental configurations. A paired-block statistical analysis identifies ResNeXtAudio and 20-second window duration as the most influential factors for performance. ERB and PCEN spectrograms significantly outperform LMS in mean AUROC, while other parameters such as stride, FFT size, and number of frequency bands show limited effect. A Grad-CAM interpretability study suggests that there are class-conditional frequency saliency differences, with AD predictions relying more on low-frequency cues. The findings provide robust design recommendations for spectrogram-based AD detection pipelines and establish a reproducible evaluation framework for future research.

Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder characterized by inattention, hyperactivity, and impulsivity. Current diagnostic methods rely primarily on subjective clinical evaluations, which are prone to bias. Neurophysiological techniques such as electroencephalography (EEG), eye tracking, and electrodermal activity (EDA) offer promising objective alternatives; however, their adoption is limited by the scarcity of large, public, multimodal datasets. To address this gap, we introduce the BALLADEER ADHD Dataset, a comprehensive multimodal resource that integrates simultaneous EEG, eye-tracking, and physiological signals from children and adolescents with ADHD and neurotypical controls. Data were collected through carefully designed cognitive tasks aimed at eliciting neurophysiological responses related to attentional control, response inhibition, and cognitive flexibility-key domains affected in ADHD. The dataset facilitates the development of machine learning models for ADHD classification and biomarker discovery through cross-modal analyses of EEG, eye movements, and autonomic nervous system activity. By publicly releasing this dataset, we aim to enhance transparency, reproducibility, and innovation in computational neuroscience and ADHD research.

BACKGROUND Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental disorder in children, characterized by inattention, hyperactivity, and impulsivity that significantly disrupt daily functioning. Comprehensive ADHD assessment requires multimethod and multi-informant evaluation, and objective attention measures can complement this process. Serious game–based assessments may further support this process by combining structured task mechanics and dynamic audiovisual guidance in a child-oriented digital environment. OBJECTIVE This study aimed to develop and validate Attention Robots, a serious game designed for assessing behavioral attention indicators in children and adolescents. The goal was to examine whether its outputs corresponded with indicators from the widely used D2-R Test of Attention as preliminary evidence of convergent validity. METHODS In this preliminary cross-sectional convergent validation study, 58 participants (34 male and 24 female individuals) aged 6-18 years, recruited through clinical referral and community sources, completed Attention Robots followed by the D2-R test. The original group variable, assigned by the project psychologists, identified 19 participants in the experimental group and 39 participants in the control group. Attention Robots presents timed rows of robot stimuli and asks participants to identify and select predefined target robots. Spearman and Pearson correlations compared working speed, omissions, commissions, concentration, and precision across both instruments. RESULTS Primary cross-test correlations were positive. Spearman correlations were observed for working speed (ρ=0.905, 95% CI 0.84-0.94; P<.001), omissions (ρ=0.333, 95% CI 0.08-0.54; P=.01), commissions (ρ=0.496, 95% CI 0.27-0.67; P<.001), concentration (ρ=0.926, 95% CI 0.88-0.96; P<.001), and precision (ρ=0.607, 95% CI 0.41-0.75; P<.001). Pearson correlations showed a similar pattern for working speed (r=0.911, 95% CI 0.85-0.95; P<.001), omissions (r=0.278, 95% CI 0.02-0.50; P=.04), commissions (r=0.396, 95% CI 0.15-0.59; P=.002), concentration (r=0.922, 95% CI 0.87-0.95; P<.001), and precision (r=0.574, 95% CI 0.37-0.72; P<.001). These findings support preliminary convergent correspondence between corresponding behavioral indicators, but they do not establish interchangeability, discriminative validity, diagnostic use, or superior engagement or acceptability. CONCLUSIONS Attention Robots provides preliminary evidence of convergent validity with the D2-R for behavioral attention indicators in children and adolescents. The innovative aspect of this study is the translation of D2-R–based attentional demands into a child-oriented serious game environment with robot-based stimuli, dynamic audiovisual guidance, and structured behavioral data capture. Unlike previous digital or serious game approaches, this study empirically compares game-derived indicators with corresponding indicators from an established attention test rather than evaluating only task feasibility, engagement, or intervention effects. This work contributes to the field by providing a preliminary validation of a serious game format that may support the collection of objective attention-related data in children and adolescents. In applied contexts, Attention Robots could be further developed as a complementary tool within multimethod ADHD assessment, although larger studies are required to evaluate subgroup performance, agreement, diagnostic use, and user experience before clinical adoption. CLINICALTRIAL

Background: Attention-Deficit Hyperactivity Disorder (ADHD) is one of the mostcommon neuro-developmental disorders diagnosed in childhood. The medical diagnosis of ADHDtypically involves extensive time-consuming interviews and tests. Electroencephalography (EEG) is amethodology to record brain activity. In recent years, several Machine Learning (ML) models havebeen developed for diagnosing ADHD using EEG data. However, they are unreliable for two reasons: 1)Performing manual pre-processing and feature selection are error-prone, and 2) training and testingthe model using the same subjects is defective. The main goal of this study is to develop a morerobust and reliable learning model for detecting ADHD. Methods: To achieve this objective, we developed an effective Deep-Learning MultiscaleConvolutional Neural Network called EEG-MSCNet. We used a Leave-One-Out subject-cross-validation approach validation to obtain more reliable results. Results: Our EEG-MSCNet model achieved an accuracy of 88.42%, surpassing the previous state-of-the-art models, including EEGNet with an accuracy of 83%. It features robust, automatic feature extraction directly from raw EEG signals. To support transparency and reproducibility, we have made the source code, databases, and logs publicly available, facilitating validation and further research. Conclusion: The proposed model achieved the best f1-score and accuracy compared withstate-of-the-art DL models. Second, an automatic end-to-end classification process could beperformed using our proposal. Consequently, it can be applied on an outpatient basis, allowingphysicians to follow their patients more closely and, as a result, speed up their treatment time.