Search Results (40)

In this paper, we first present a multimodal architecture called CrossAttn-v1. This model is designed to recover punctuation marks in Kazakh and combines contextual XLM-RoBERTa-large text embeddings with the Whisper large-v3 encoder states via a cross-attention mechanism. In addition, a 4-dimensional prosodic vector and a CRF output layer are used. The model was trained using an adapted Whisper ASR model on 33,332 utterances from the KazakhTTS2 corpus. After adaptation, the word error rate decreased from 45.7% to 4.25%. On the in-domain test set (56,396 tokens), CrossAttn-v1 achieved F1-macro = 0.8485 for recovering five-class punctuation marks. Furthermore, CrossAttn-v1 outperformed the GPT-4o zero-shot model by +0.294 F1 and the M3 Hybrid model based on prosody alone by +0.070 F1. The class analysis showed that the Whisper encoder states were particularly useful for prosody-dependent punctuation. For example, it outperformed M3 Hybrid by +9.5 percentage points on the QUESTION mark and by +20.2 percentage points on the EXCLAIM mark. On 883 out-of-domain natural speech recordings, the model performed similarly to the text-only baseline model (Δ = −0.041, not significant), suggesting that domain mismatch in the Whisper training corpus was a major factor limiting generalization. Full article

Background/Objectives: Attention-deficit hyperactivity disorder (ADHD) in adults is often overlooked in refugees, especially those displaced by war and diagnosed with chronic issues such as hypertension and type 2 diabetes mellitus. We sought to provide a preliminary screening to align with ADHD screening and to examine its related demographic and clinical factors. Methods: This cross-sectional study recruited Syrian refugees residing in Jordan. The Adult ADHD Self-Report Scale-V1.1 (ASRS) was employed to provide a preliminary screening of “Consistency with ADHD screening”. Multivariable regression analysis was used to identify the risk factors associated with the “Consistency with ADHD screening”. Results: Data analysis included 345 patients; 179 (51.9%) were females. β-Blockers were used in 151 (43.8%), metformin in 134 (38.8%), and sulfonylurea in 86 (24.9%). In the study, 158 participants (45.8%) reported very consistent results related to ADHD. We used multivariate binary logistic regression, which showed that in our groups that received metformin (OR of 2.08, 95% CI 1.32–3.27, p = 0.001) and insulin (OR of 2.25, 95% CI 1.00–5.04, p = 0.04), we observed a positive association with ADHD symptoms. Also, we noted that high school education was negatively associated with the results of the ADHD screen (OR of 0.58, 95% CI of 0.36–0.94, p = 0.02). Conclusions: This preliminary study guides future steps in addressing ADHD symptoms in war-displaced refugees; proper diabetes management and education seem to be important factors. Full article

This study rigorously evaluates 13 Convolutional Neural Network (CNN) architectures for Thai dialect recognition. By treating Automatic Speech Recognition (ASR) as a computer vision texture classification task, we processed an extensive 840-h dataset from the Spoken Language Systems, Chulalongkorn University (SLSCU) corpus. Raw audio from four major dialects—Central, Northern (Khummuang), Northeastern (Korat), and Southern (Pat-tani)—was transformed into 2D Mel-spectrograms using the Short-Time Fourier Transform (STFT). We analyzed a diverse range of architectures, including the VGG, Inception, ResNet, DenseNet, and MobileNet families, to establish the optimal trade-off between mathematical complexity and spectral feature extraction. Our experimental results identify NASNet-Mobile as the most effective model, achieving a macro-average F1-score of 0.9425. The analysis suggests that NASNet’s search-optimized cell structure is uniquely capable of capturing the multiscale texture of phonetic formants. In contrast, we observed a catastrophic mode collapse in VGG16 (32.97% accuracy), likely due to excessive parameter bloat, while Xception and MobileNetV2 maintained robust generalization. Confusion matrix analysis reveals high acoustic distinctiveness for Southern Thai (96.7% recall), whereas Northern Thai exhibits significant spectral overlap with Central Thai. These results support the hypothesis that CNNs interpret spectrograms as textures rather than discrete objects, positioning NASNet-Mobile as a high-performance, low-latency baseline for edge-device deployment in resource-constrained environments. Full article

For Autonomous Vehicles (AVs), recognizing traffic lights and signs is critical for safety because perception errors directly affect navigation decisions. Real-world disturbances such as glare, rain, dirt, and graffiti, as well as digital adversarial attacks, can lead to dangerous misclassifications. Current research lacks (i) temporal continuity (stable detection across consecutive frames to prevent flickering misclassifications), (ii) multi-field-of-view (FoV) sensing, and (iii) integrated defenses against both digital and natural degradation. This paper presents two principal contributions: (1) a three-layer defense framework integrating feature squeezing, inference-time temperature scaling (softmax $\tau$ = 3 without distillation training), and entropy-based anomaly detection with sequence-level temporal voting; (2) a 500 sequence dual-FoV benchmark (30k base frames, 150k with perturbations) from aiMotive, Waymo, Udacity, and Texas sources across four operational design domains. The unified defense stack achieves 79.8% mAP on a 100-sequence test set (6k base frames, 30k with perturbations), reducing attack success rate from 37.4% to 18.2% (51% reduction) and high-risk misclassifications by 32%. Cross-FoV validation and temporal voting enhance stability under lighting changes (+3.5% mAP) and occlusions (+2.7% mAP). Defense improvements (+9.5–9.6% mAP) remain consistent across native 3D (aiMotive, Waymo) and projected 2D (Udacity, Texas) annotations. Preliminary recapture experiments (n = 15 scenarios) show 2.5% synthetic–physical ASR gap (p = 0.18), though larger validation is needed. Code, models, and dataset reconstruction tools are publicly available. Full article

Background and Objectives: The dopamine D2 receptor (DRD2) plays a central role in fronto-striatal circuits regulating cognitive control and reward processing. The rs1076560 polymorphism alters receptor isoform expression, potentially modifying impulsivity and vulnerability to stimulant use disorders. We examined gene–environment interactions between rs1076560 and stimulant dependence in relation to impulsivity, ADHD traits, and hedonic capacity. Methods: A total of 517 men (235 stimulant-dependent, 282 controls) completed the Barratt Impulsiveness Scale (BIS-11), Adult ADHD Self-Report Scale (ASRS v1.1), and Snaith–Hamilton Pleasure Scale (SHAPS). Genotyping for rs1076560 was performed using real-time PCR, and two-way ANOVAs tested genotype-by-group effects. Results: Significant genotype-by-group interactions were observed across all BIS-11 subscales and ASRS scores. In the stimulant-dependent group, C/C homozygotes showed the highest levels of attentional impulsivity and attentional dysregulation compared to both A/C and C/C controls. In contrast, within the control group, A/A homozygotes demonstrated higher motor impulsivity, non-planning impulsivity, and BIS-11 total scores than C/C controls. No significant main effects or interactions were found for SHAPS scores. Conclusions: DRD2 rs1076560 moderates impulsivity-related traits through dopaminergic pathways relevant to executive dysfunction in stimulant use disorders. These findings highlight a neurobiological mechanism of addiction vulnerability and may inform precision approaches in neurology and psychiatry. Full article

Background: This study examined the impact of aerobic capacity on force–velocity (F–v) variables and repeated-sprint (RS) performance in male national-level sprinters (SPRs, n = 8; 177.0 ± 4.3 cm; 74.0 ± 5.0 kg; maximal oxygen uptake [VO₂max]: 55.4 ± 3.0 mL/kg/min) and middle-distance runners (MDRs; n = 8; 179.0 ± 5.1 cm; 67.2 ± 5.0 kg; VO₂max: 64.3 ± 3.3 mL/kg/min). Method: Participants underwent assessments of aerobic capacity, mechanical F-v profiling in sprinting 2 × 60 m with full recovery, and a 10 × 60 m repeated-sprint test with 30 s recovery. Results: MDRs exhibited significantly higher VO₂max (p < 0.001) and speed at VO₂max (vVO₂max, p < 0.001), while SPRs demonstrated greater anaerobic speed reserve (ASR, p < 0.001), maximal theoretical horizontal force (F₀, p = 0.012), and power output (P_max, p < 0.01). During the RS test, SPRs displayed a 16.6% performance decrement (p = 0.002) and failed to complete all sprints with voluntary withdrawal after 5–8 sprints due to exhaustion, whereas MDRs maintained consistent performance. SPRs exhibited a larger decrease in v₀ compared to MDRs (p < 0.01), whereas no differences were observed on F₀ (p = 0.519) and P_max (p = 0.758). Blood lactate accumulation was higher in SPRs (p < 0.001). Multiple linear regression analysis on the pooled sample identified vVO₂max (p = 0.003) and not ASR (p = 0.482) as a key predictor of fatigue resistance. Conclusions: These findings underscore the critical role of aerobic capacity in sustaining RS performance. Aerobic capacity, specifically vVO₂max, emerged as the primary determinant of fatigue resistance during repeated sprints, underscoring its critical role in sustaining RS performance over mechanical variables such as v₀ but not F₀ and P_max. Full article

Background: Attention-deficit/hyperactivity disorder (ADHD) is a chronic neurodevelopmental disorder that continues into adulthood and is linked to significant academic, occupational, and psychological challenges. Medical students may be at heightened risk due to the cognitive and emotional demands of their training. However, regional data on adult ADHD symptoms in this population, particularly in northern Saudi Arabia, remain limited. Objective: to estimate the prevalence of probable adult ADHD symptoms among medical students at the University of Tabuk and to examine the associated sociodemographic, academic, and health-related factors. Methods: A cross-sectional, web-based survey was conducted between 15 May and 5 June 2025 among randomly selected Saudi medical students (years 2–6) at the University of Tabuk. The validated Arabic version of the WHO Adult ADHD Self-Report Scale (ASRS-v1.1) was used to screen for probable ADHD. A positive screen was defined as ≥4 flagged items. Descriptive and bivariate analyses were performed using SPSS v29. Results: Of the 219 respondents (60.3% male; mean age: 21.6 years), 23.3% screened positive for probable adult ADHD. Symptom frequency peaked in the third (40.0%) and fourth (35.6%) academic years and was lowest among sixth-year students (11.4%) (p = 0.012). A strong association was observed between positive ADHD screening and self-reported psychiatric disorders (p < 0.001). No statistically significant associations were found for gender, income, GPA, marital status, or academic phase. Conclusions: Nearly one in four medical students at the University of Tabuk exhibited symptoms suggestive of adult ADHD, a prevalence markedly higher than global estimates and consistent with regional trends. The association with psychiatric morbidity and the mid-programme peak suggests a need for proactive screening, mental health support, and academic accommodations. Universities can translate these findings into practice by instituting routine ADHD screening, offering flexible assessment accommodations, embedding peer-mentoring programmes, and strengthening on-campus mental-health referral pathways. Full article

In the field of water-based inks, the use of alkali-soluble resins (ASRs) as polymeric surfactants for synthesizing polyacrylate latexes has become a mainstream method. This study first designed and prepared crosslinkable ASRs with a diacetone acrylamide (DAAM) crosslinking monomer via emulsion polymerization. These ASRs were then employed as surfactants to synthesize self-crosslinking polyacrylate latexes through an in situ one-pot method, systematically investigating the influence of crosslinkable ASRs on the properties of the corresponding polyacrylate latexes. The research revealed that all prepared polyacrylate latexes exhibited a core–shell structure. With increasing DAAM content in the ASRs, the latex particle size gradually increased while the particle size distribution narrowed. All latexes demonstrated excellent stability, with absolute ζ-potential values exceeding 30 mV. The introduction of DAAM into ASRs significantly increased the glass transition temperature in the high-temperature region of the corresponding latex films, with the tensile strength reaching a maximum of 7.96 MPa. Moderate crosslinking in ASRs substantially improved the water resistance of latex films. Crosslinking degree tests indicated that latex films prepared through either single shell-layer crosslinking or single core-layer crosslinking showed relatively low crosslinking degrees, while only the dual core–shell crosslinking strategy could effectively enhance the film crosslinking degree. However, excessively crosslinked shell layers significantly hindered the crosslinking reaction of DAAM in the core layer, leading to reduced overall film crosslinking. Additionally, incorporating a certain number of DAAM crosslinking groups in ASRs was found to improve the adhesion of corresponding water-based inks on PE and BOPP substrates, with adhesion on BOPP substrates reaching up to 100%. Full article

Objective: to assess the association between attention-deficit/hyperactivity disorder (ADHD), type 1 diabetes mellitus (T1D), and cardiovascular comorbidities in adults. Methods: The Adult Self-Report Scale V1.1 (ASRS) for ADHD symptoms was electronically sent to 2069 adults with T1D. Cardiometabolic conditions, laboratory measurements, and PHQ-2/PHQ-9 depression scores were obtained from the electronic medical record. Results: In total, 292 (14.1%) individuals responded and 279 consented to medical records extraction. The average age was 47.4 years (SD: ±18.9), 64.2% were women, 95.7% were non-Hispanic white, and the mean HbA1c level was 7.7% (±1.5%). Of 273 completing ASRS, 87 med ADHD criteria (ASRS positive, 31.9%), and 42 (15.4%) had an ADHD diagnosis or medication. Women had higher scores than men. ADHD symptoms decreased with age, but remained significantly higher than the general population levels. HbA1c levels were positively associated with the ASRS scores (Spearman’s r = 0.28, p < 0.0001). ASRS positive individuals had worse glycemic control (HbA1c ≥ 8.0%, adjusted OR 2.3, 95%CI: 1.3–4.1, p < 0.0001) and higher PHQ-9 scores (10 ± 7.3 vs. 6.1 ± 6, χ²₍₁₎ = 9.2, p = 0.002) than the ASRS negative group. No associations were found between ASRS scores and cardiometabolic diseases, or other laboratory or clinical measurements. Conclusions: Many adults with T1D exhibit undiagnosed ADHD symptoms, which correlate with poorer glycemic control and depression. Further research with larger samples is needed to investigate ADHD prevalence and impacts in this group. Full article

Automatic Speech Recognition (ASR) systems have made consistent advancements, achieving notable improvements in state-of-the-art performance across various languages. However, their effectiveness often declines significantly in low-resource settings, where data and linguistic resources are limited. This paper addresses the challenges of ASR for a low-resource language, Khalkha Mongolian, by leveraging a transliteration-aided transfer learning approach. Specifically, it improves the ASR system for Khalkha Mongolian by transliterating text from a well-resourced Chakhar Mongolian (Uighur script) dataset to the Cyrillic script and then fine-tuning it with Khalkha Mongolian data. The method effectively enhances the ASR performance of Khalkha Mongolian. The effectiveness of the proposed method was validated on three popular ASR models, Wav2Vec2-BERT, Conformer-Large, and Whisper-large-v3. Among these models, the best relative improvement in word error rate (WER) reaches 32.50%, while the absolute improvement reaches 19.26%. Full article

Background/Objectives: Iron deficiency has been suggested as a potential mechanism for attention-deficit hyperactivity disorder (ADHD) development due to involvement in neurotransmitter synthesis and transporter expression. As iron deficiency is particularly common in women of reproductive age, often due to heavy menstrual bleeding (HMB), we aimed to explore the relationship between iron deficiency, HMB and ADHD in women. Methods: We screened women (18–49 years) at university and local sporting events in Western Australia. To screen for ADHD, section A of the Adult ADHD Self-Report Scale-V1.1 (ASRS-V1.1) and the Adult Concentration Inventory were used to assess cognitive disengagement syndrome (CDS) symptoms. Risk factors for iron deficiency, such as HMB, commonly reported symptoms and a fingerpick haemoglobin concentration (Hb) (Hemocue Hb801) were recorded. Results: Of the 405 completed questionnaires, the mean age was 24.8 ± 10.1 years, the mean Hb was 136.8 ± 12.4 g/L and 6.4% of women were anaemic. Symptoms suggestive of ADHD were reported by 174/405 (43%) women, and 128/405 (32%) women reported HMB. There was a greater prevalence of HMB reported in those experiencing symptoms suggestive of ADHD (39% vs. 26%, p = 0.01). Symptoms of fatigue, dizziness, brain fog, anxiety, heart palpitations, headaches, restless legs and depression were more common in patients with symptoms suggestive of ADHD (p ≤ 0.01) and HMB (p < 0.05). Anaemia status did not influence ADHD status (p = 0.87) nor CDS scores (15.7 ± 7.0 vs. 13.8 ± 6.1, p = 0.17). Conclusions: There is an apparent relationship between those with symptoms reported in ADHD, HMB and iron deficiency. Further exploration is required to determine whether there is a causative relationship. Full article

Objectives: The treatment of emotional dysregulation (ED) poses a major challenge for clinicians managing adult attention-deficit/hyperactivity disorder (ADHD). This naturalistic longitudinal study aimed to evaluate the effects of combining mood stabilizers (MS) with standard pharmacotherapy in this population. Methods: Fifty-six adult patients with ADHD, with or without bipolar spectrum disorders, who were followed-up for at least 4 months at Pisa University Hospital were included and grouped based on the prescription of ADHD treatment with prior MS, with conomitant MS and without MS. Changes in self-reported ED, self-reported and informant-reported ADHD severity were assessed using RIPoSt-40, ASRS-v1.1, and CAARS-O:SV. Longitudinal analyses were conducted separately for each group using a pairwise one-sample paired Student’s t-test. Results: A significant reduction in ED severity was observed in those treated with methylphenidate (MPH) and concomitant MS and in those with atomoxetine (ATX) without MS. Negative emotionality and emotional impulsivity significantly decreased in both these groups, while affective instability only improved in those with MPH and concomitant MS. Self-reported ADHD improvements were significant in all groups receiving MPH, whether with concomitant, prior, or without MS. Significant changes in informant-reported ADHD severity were found in those receiving MPH with concomitant or prior MS. Conclusions: The findings highlight the benefits of concomitant MS and MPH treatment on ED, suggest a preferential effect of ATX on negative emotionality, and confirm the effectiveness of MPH for adult ADHD symptoms, regardless of additional treatment with MS. Further studies are needed to explore whether and how MS and MPH may complement each other in reducing ED. Full article

Introduction: In 2019, an updated version of the Diagnostic Interview for ADHD in Adults (DIVA-5) was developed based on DSM-5 criteria, currently validated in Korean and Farsi. The aim of this study is to validate the DIVA-5 Italian version. Methods: 132 subjects in the Adult ADHD Screening Center of AUSL-Modena, who agreed to participate in this study, were selected. Socio-demographic and clinical variables were collected. DIVA-5, Barkley Adult ADHD Rating Scale (BAARS), and Adult ADHD Self Rating Scale (ASRS-v1.1) were administered. We assessed the internal consistency of the DIVA-5 Italian version and its concurrent validity with ASRS-v1.1 and BAARS-IV. An exploratory factor analysis (EFA) was conducted to evaluate the construct validity, and a multiple linear regression to evaluate the predictive validity. Results: Our analysis indicated good internal consistence of the DIVA-5 Italian version (Cronbach’s alpha and Kuder coefficients ranged between 0.61 and 0.78). The EFA showed five factors representing specific variance. The correlation between the corresponding ADHD dimensions of DIVA-5 and BAARS was found to be statistically significant (Spearman’s coefficient ranged between 0.61 and 0.47, p = 0.000), while the correlation between the DIVA-5 dimensions and ASRS-v1.1 was statistically significant for all the dimensions except child hyperactivity/impulsivity. The multiple linear regression showed a positive association of the DIVA-5 score with the “job” variable and a negative association with “drug therapy”. DIVA-5 showed greater sensitivity for inattention in adulthood and greater specificity for hyperactivity/impulsivity in childhood. Conclusions: Our results confirm that the DIVA-5 Italian version represents a valid and reliable tool to diagnose adult ADHD. Full article

Dregs waste, a byproduct from green liquor clarification in the pulp industry, is increasingly generated as global cellulose production rises. This accumulation, along with its underutilization, presents environmental challenges and opportunities for reuse. The study focuses on the high alkali content in dregs, which can impact cement durability. The main objective is to analyze the effect of dregs on the alkali–silica reaction in mortars. Dregs were incorporated into mortar mixtures at 0%, 5%, 10%, and 15% proportions relative to cement mass, using six types of Brazilian cement and a blend with silica fume. The alkali–silica reactivity was assessed via the accelerated mortar prism test (ABNT NBR 15577-4:2018), with compressive strength tests and scanning electron microscopy evaluating structural integrity and microstructural changes. The results indicated that adding 5% dregs improved compressive strength in certain mortars, such as CP V-ARI RS, CP II-F, and CP IV. However, at 15% dregs, compressive strength significantly decreased, particularly in CP V with 10% silica fume. Cements with high pozzolanic content, such as CP IV and CP III, showed strong potential to inhibit ASR expansion. However, of the 28 mixtures analyzed, only four containing CP III had expansions within the limits set by standards. This study highlights the potential of incorporating dregs as a supplementary material in cement, promoting sustainability in the industry and reducing environmental impact. Full article

Background: The relationship between adult ADHD symptoms in People with Multiple Sclerosis (PwMS) is understudied. This study aimed to answer two questions: are PwMS more likely to experience higher ADHD symptoms versus healthy subjects? And what are the correlates of severe ADHD symptoms in PwMS? Methods: This study followed a cross-sectional design with predefined inclusion criteria. The Adult ADHD Self-Report Scale-V1.1 (ASRS) was used to assess the ADHD symptoms severity. Results: Data were analyzed from 171 PwMS and 200 controls. Regression analysis revealed that PwMS were at a significantly (B = 3.05, t = 2.24, 95% CI = 0.37–5.73, p = 0.02) higher risk to report higher ADHD scores versus controls. In addition, PwMS with relapses in the last 6 months and PwMS reporting smartphone addiction were at a significantly higher risk for severe ADHD (B = 7.19, t = 269, 95% CI = 1.91–12.48, p = 0.008) and (B = 9.18, t = 3.47, 95% CI = 3.97–14.41, p = 0.001), respectively. In conclusion, diagnosis with MS in our study was identified as a risk for higher ADHD symptoms. Conclusions: Further research is required to establish this relationship, and holistic medical and psychological interventions are required to improve the cognitive status of PwMS. Full article