Search Results (36)

Background/Objectives: Schizophrenia Spectrum Disorders (SSDs) carry a debilitating burden of disease which, even after pharmacological and psychological treatment are optimized, remains difficult to fully target. New online-delivered and user-led interventions may provide an appropriate, cost-effective answer to this problem. This study aims to retrieve the currently gathered findings on the efficacy of these interventions across several outcomes, such as symptom severity, social cognition, functioning and others. Methods: A systematic review of the current available literature was conducted. Of 29 potentially relevant articles, 26 were included and assigned at least one of four intervention types: Web-Based Therapy (WBT), Web-Based Psycho-Education (WBP), Online Peer Support (OPS) and Prompt-Based Intervention (PBI). Results: The findings were grouped based on outcome. Of 24 studies evaluating the effects of symptom severity, 14 have achieved statistically significant results, and 10 have not. WBT (such as online-delivered Cognitive Behavioral Therapy, Acceptance and Commitment Therapy, social cognition training and Mindfulness Training) seemed to be the most effective at targeting symptoms. Of 14 studies evaluating functioning, seven achieved significant results, four involving a form of social or neurocognitive training, suggesting a potential pathway towards functional improvements through interventions targeting cognition and motivation. Regarding social cognition, all seven studies measuring the effects of an intervention on this outcome produced significant results, indicating that this outcome lends itself well to remote, online administration. This may be linked with the nature of social cognition exercises, as they are commonly administered through a digital medium (such as pictures, videos and auditory exercises), a delivery method that suits the online-user led model very well. Conclusions: Online user-led interventions show promise as a new way to tackle functional deficits in SSD patients and achieve these improvements through targeting social cognition, a hard-to-reach component of the burden of SSDs which seems to be successfully targetable in a remote, user-led fashion. Symptomatic improvements can also be achievable, through the combination of these interventions with treatment as usual. Full article

Background: Out-of-hospital cardiac arrest (OHCA) survivors and their relatives may face challenges following hospital discharge, relating to mood, cognition, and returning to normal day-to-day activities. Identified research gaps include a lack of knowledge around what type of intervention is needed to best navigate recovery. In this study, we investigate the feasibility and patient acceptability of a new virtual psychoeducational group intervention for OHCA survivors and their relatives and compare it to a control group receiving a digital information booklet. Methods: V-CARE is a comparative, single-blind randomized pilot trial including participants at selected sites of the STEPCARE trial, in the United Kingdom and Sweden. Inclusion criteria are a modified Rankin Scale (mRS) ≤ 3 at 30-day follow-up; no diagnosis of dementia; and not experiencing an acute psychiatric episode. One caregiver per patient is invited to participate optionally. The intervention group in V-CARE receives four semi-structured, one-hour-long, psychoeducational sessions delivered remotely via video call by a trained clinician once a week, 2–3 months after hospital discharge. The sessions cover understanding cardiac arrest; coping with fatigue and memory problems; managing low mood and anxiety; and returning to daily life. The control group receives an information booklet focused on fatigue, memory/cognitive problems, mental health, and practical coping strategies. Results: Primary: feasibility (number of patients consented) and acceptability (retention rate); secondary: satisfaction with care (Client Satisfaction Questionnaire 8 item), self-management skills (Self-Management Assessment Scale) and, where available, health-related outcomes assessed in the STEPCARE Extended Follow-up sub-study including cognition, fatigue, mood, quality of life, and return to work. Conclusions: If preliminary insights from the V-CARE trial suggest the intervention to be feasible and acceptable, the results will be used to design a larger trial aimed at informing future interventions to support OHCA recovery. Full article

Background/Objectives: Sleep disorders are prevalent among aging populations and are often linked to cognitive decline, chronic conditions, and reduced quality of life. Traditional diagnostic methods, such as polysomnography, are resource-intensive and limited in accessibility. Meanwhile, individuals frequently describe their sleep experiences through unstructured narratives in clinical notes, online forums, and telehealth platforms. This study proposes a machine learning pipeline (sleepCare) that classifies sleep-related narratives into clinically meaningful categories, including stress-related, neurodegenerative, and breathing-related disorders. The proposed framework employs natural language processing (NLP) and machine learning techniques to support remote applications and real-time patient monitoring, offering a scalable solution for the early identification of sleep disturbances. Methods: The sleepCare consists of a three-tiered classification pipeline to analyze narrative sleep reports. First, a baseline model used a Multinomial Naïve Bayes classifier with n-gram features from a Bag-of-Words representation. Next, a Support Vector Machine (SVM) was trained on GloVe-based word embeddings to capture semantic context. Finally, a transformer-based model (BERT) was fine-tuned to extract contextual embeddings, using the [CLS] token as input for SVM classification. Each model was evaluated using stratified train-test splits and 10-fold cross-validation. Hyperparameter tuning via GridSearchCV optimized performance. The dataset contained 475 labeled sleep narratives, classified into five etiological categories relevant for clinical interpretation. Results: The transformer-based model utilizing BERT embeddings and an optimized Support Vector Machine classifier achieved an overall accuracy of 81% on the test set. Class-wise F1-scores ranged from 0.72 to 0.91, with the highest performance observed in classifying normal or improved sleep (F1 = 0.91). The macro average F1-score was 0.78, indicating balanced performance across all categories. GridSearchCV identified the optimal SVM parameters (C = 4, kernel = ‘rbf’, gamma = 0.01, degree = 2, class_weight = ‘balanced’). The confusion matrix revealed robust classification with limited misclassifications, particularly between overlapping symptom categories such as stress-related and neurodegenerative sleep disturbances. Conclusions: Unlike generic large language model applications, our approach emphasizes the personalized identification of sleep symptomatology through targeted classification of the narrative input. By integrating structured learning with contextual embeddings, the framework offers a clinically meaningful, scalable solution for early detection and differentiation of sleep disorders in diverse, real-world, and remote settings. Full article

Background/Objectives: Rett Syndrome (RTT) is a rare neurodevelopmental disorder that affects girls and is characterized by severe motor and cognitive impairments, the loss of purposeful hand use, and communication difficulties. Children with RTT, especially those aged 5 to 9 years, often struggle to develop the foundational skills necessary for adaptive functioning, such as eye contact, object tracking, functional gestures, turn-taking, and basic communication. These abilities are essential for cognitive, social, and motor development and contribute to greater autonomy in daily life. This study aimed to explore the feasibility of a structured telerehabilitation program and to provide preliminary observations of its potential utility for young girls with RTT, addressing the presumed challenge of engaging this population in video-based interactive training. Methods: The intervention consisted of 30 remotely delivered sessions (each lasting 90 min), with assessments at baseline (A), after 5 weeks (B1), and after 10 weeks (B2). Quantitative outcome measures focused on changes in eye contact, object tracking, functional gestures, social engagement, and responsiveness to visual stimulus. Results: The findings indicate that the program was feasible and well-tolerated. Improvements were observed across all measured domains, and participants showed high levels of engagement and participation throughout the intervention. While these results are preliminary, they suggest that interactive digital formats may be promising for supporting foundational learning processes in children with RTT. Conclusions: This study provides initial evidence that telerehabilitation is a feasible approach for engaging young girls with RTT and supporting adaptive skill development. These findings may inform future research and the design of controlled studies to evaluate the efficacy of technology-assisted interventions in this population. Full article

Background/Objectives: Alzheimer’s disease (AD) is a progressive, neurodegenerative disorder, which causes memory loss and loss of cognitive functioning, along with behavioral changes. Early detection is important to delay disease progression, timely intervention and to increase patients’ and caregivers’ quality of life (QoL). One of the major and primary challenges for preventing any disease is to identify the disease at the initial stage through a quick and reliable detection process. Different researchers across the world are still working relentlessly, coming up with significant solutions. Artificial intelligence-based solutions are putting great importance on identifying the disease efficiently, where deep learning with medical imaging is highly being utilized to develop disease detection frameworks. In this work, a novel and optimized detection framework has been proposed that comes with remarkable performance that can classify the level of Alzheimer’s accurately and efficiently. Methods: A powerful vision transformer model (ViT-B16) with three efficient Convolutional Neural Network (CNN) models (VGG19, ResNet152V2, and EfficientNetV2B3) has been trained with a benchmark dataset, ‘OASIS’, that comes with a high volume of brain Magnetic Resonance Images (MRI). Results: A weighted average ensemble technique with a Grasshopper optimization algorithm has been designed and utilized to ensure maximum performance with high accuracy of 97.31%, precision of 97.32, recall of 97.35, and F1 score of 0.97. Conclusions: The work has been compared with other existing state-of-the-art techniques, where it comes with high efficiency, sensitivity, and reliability. The framework can be utilized in IoMT infrastructure where one can access smart and remote diagnosis services. Full article

Non-Intrusive Load Monitoring (NILM) includes a set of methods orientated to disaggregating the power consumption of a household per appliance. It is commonly based on a single metering point, typically a smart meter at the entry of the electrical grid of the building, where signals of interest, such as voltage or current, can be measured and analyzed in order to disaggregate and identify which appliance is turned on/off at any time. Although this information is key for further applications linked to energy efficiency and management, it may also be applied to social and health contexts. Since the activation of the appliances in a household is related to certain daily activities carried out by the corresponding tenants, NILM techniques are also interesting in the design of remote monitoring systems that can enhance the development of novel feasible healthcare models. Therefore, these techniques may foster the independent living of elderly and/or cognitively impaired people in their own homes, while relatives and caregivers may have access to additional information about a person’s routines. In this context, this work describes an intelligent solution based on deep neural networks, which is able to identify the daily activities carried out in a household, starting from the disaggregated consumption per appliance provided by a commercial smart meter. With the daily activities identified, the usage patterns of the appliances and the corresponding behaviour can be monitored in the long term after a training period. In this way, every new day may be assessed statistically, thus providing a score about how similar this day is to the routines learned during the training interval. The proposal has been experimentally validated by means of two commercially available smart monitors installed in real houses where tenants followed their daily routines, as well as by using the well-known database UK-DALE. Full article

Cardiovascular diseases pose a significant threat to global health, and cardiac rehabilitation (CR) has become a critical component of patient care. Heart Rate Variability Biofeedback (HRVB) is a non-invasive approach that helps modulate the Autonomic Nervous System (ANS) through Resonance Frequency (RF) breathing, supporting CR for cardiovascular patients. However, traditional HRVB techniques rely heavily on manual RF selection and face-to-face guidance, limiting their widespread application, particularly in home-based CR. To address these limitations, we propose a remote human-computer collaborative HRVB system, “FreeResp”, which features autonomous RF adjustment through a simplified cognitive computational model, eliminating the reliance on therapists. Furthermore, the system integrates wearable technology and the Internet of Things (IoT) to support remote monitoring and personalized interventions. By incorporating tactile guidance technology with an airbag, the system assists patients in performing diaphragmatic breathing more effectively. FreeResp demonstrated high consistency with conventional HRVB methods in determining RF values (22/24) from 24 valid training samples. Moreover, a one-month home-based RF breathing training using FreeResp showed significant improvements in Heart Rate Variability (HRV) (p < 0.05). These findings suggest that FreeResp is a promising solution for home-based CR, offering timely and precise interventions and providing a new approach to long-term cardiovascular health management. Full article

China’s Yellow (Bohai) Sea bird habitat is an important ecological region. Its unique ecology and challenges provide rich resources for research and study. Our course design concept is supported by AI technology, and improves students’ abilities through innovative functions such as dynamic data support, personalized learning paths, immersive research and study experience, and diversified evaluation mechanisms. The course content revolves around the “human–land coordination concept”, including pre-trip thinking, research and study during the trip, and post-trip exhibition learning, covering regional cognition, remote sensing image analysis, field investigation, and protection plan display activities. ERNIE Bot participates in optimizing the learning path throughout the process. The course evaluation system starts from the three dimensions of “land to people”, “people to land”, and the “coordination of the human–land relationship”, adopts processes and final evaluation, and uses ERNIE Bot to achieve real-time monitoring, data analysis, personalized reports, and dynamic feedback, improving the objectivity and efficiency of evaluation, and helping students and teachers optimize learning and teaching. However, AI has limitations in geographical research and study, such as insufficient technical adaptability, the influence of students’ abilities and habits, and the adaptation of teachers’ role changes. To this end, optimization strategies such as improving data quality and technical platforms, strengthening student technical training, enhancing teachers’ AI application capabilities, and enriching AI functions and teaching scenarios are proposed to enhance the application effect of AI in geographical research and promote innovation in educational models and student capacity building. Full article

Remote sensing and computer vision technologies are increasingly leveraged for rapid post-disaster building damage assessment, becoming a crucial and practical approach. In this context, the accuracy of employing various AI models in pixel-level change detection methods is significantly dependent on the consistency between pre- and post-disaster building images, particularly regarding variations in resolution, viewing angle, and lighting conditions; in object-level feature recognition methods, the low richness of semantic details of damaged buildings in images leads to a poor detection accuracy. This paper proposes a novel method, OCD-BDA (Object-Level Change Detection for Post-Disaster Building Damage Assessment), as an alternative to pixel-level change detection and object-level feature recognition methods. Inspired by human cognitive processes, this method incorporates three key steps: an efficient sample acquisition for object localization, labeling via HGC (Hierarchical and Gaussian Clustering), and model training and prediction for classification. Furthermore, this study establishes a change detection dataset based on Google Earth imagery of regions in Hatay Province before and after the Turkish earthquake. This dataset is characterized by pixel inconsistency and significant differences in photographic angles and lighting conditions between pre- and post-disaster images, making it a valuable test dataset for other studies. As a result, in the experiments of comparative generalization capabilities, OCD-BDA demonstrated a significant improvement, achieving an accuracy of 71%, which is twice that of the second-ranking method. Moreover, OCD-BDA exhibits superior performance in tasks with small sample amounts and rapid training time. With only 1% of the training samples, it achieves a prediction accuracy exceeding that of traditional transfer learning methods with 60% of samples. Additionally, it completes assessments across a large disaster area (450 km²) with 93% accuracy in under 23 min. Full article

Road extraction techniques based on remote sensing image have significantly advanced. Currently, fully supervised road segmentation neural networks based on remote sensing images require a significant number of densely labeled road samples, limiting their applicability in large-scale scenarios. Consequently, semi-supervised methods that utilize fewer labeled data have gained increasing attention. However, the imbalance between a small quantity of labeled data and a large volume of unlabeled data leads to local detail errors and overall cognitive mistakes in semi-supervised road extraction. To address this challenge, this paper proposes a novel consistency self-training semi-supervised method (CSSnet), which effectively learns from a limited number of labeled data samples and a large amount of unlabeled data. This method integrates self-training semi-supervised segmentation with semi-supervised classification. The semi-supervised segmentation component relies on an enhanced generative adversarial network for semantic segmentation, which significantly reduces local detail errors. The semi-supervised classification component relies on an upgraded mean-teacher network to handle overall cognitive errors. Our method exhibits excellent performance with a modest amount of labeled data. This study was validated on three separate road datasets comprising high-resolution remote sensing satellite images and UAV photographs. Experimental findings showed that our method consistently outperformed state-of-the-art semi-supervised methods and several classic fully supervised methods. Full article

Conventional patient monitoring methods require skin-to-skin contact, continuous observation, and long working shifts, causing physical and mental stress for medical professionals. Remote patient monitoring (RPM) assists healthcare workers in monitoring patients distantly using various wearable sensors, reducing stress and infection risk. RPM can be enabled by using the Digital Twins (DTs)-based Internet of Robotic Things (IoRT) that merges robotics with the Internet of Things (IoT) and creates a virtual twin (VT) that acquires sensor data from the physical twin (PT) during operation to reflect its behavior. However, manual navigation of PT causes cognitive fatigue for the operator, affecting trust dynamics, satisfaction, and task performance. Also, operating manual systems requires proper training and long-term experience. This research implements autonomous control in the DTs-based IoRT to remotely monitor patients with chronic or contagious diseases. This work extends our previous paper that required the user to manually operate the PT using its VT to collect patient data for medical inspection. The proposed decision-making algorithm enables the PT to autonomously navigate towards the patient’s room, collect and transmit health data, and return to the base station while avoiding various obstacles. Rather than manually navigating, the medical personnel direct the PT to a specific target position using the Menu buttons. The medical staff can monitor the PT and the received sensor information in the pre-built virtual environment (VE). Based on the operator’s preference, manual control of the PT is also achievable. The experimental outcomes and comparative analysis verify the efficiency of the proposed system. Full article

Cerebral palsy is a disorder of central motor and postural development, resulting in limited mobility. Cerebral palsy is often accompanied by cognitive impairment and abnormal behavior, significantly impacting individuals and society. Time, energy, and economic investment in the rehabilitation process is substantial, yet the rehabilitation outcomes often remain unsatisfactory. Additionally, some patients have limited sensory perception during rehabilitation training, making it challenging to effectively regulate exercise intensity. Traditional evaluation methods are mostly based on recovery performance, lack guidance at the neurophysiological level, and have an unequal distribution of medical rehabilitation resources, which pose great challenges to the rehabilitation of patients. Based on the issues mentioned above, this paper proposes a real-time cerebral signal monitoring system based on wearable devices. This system can monitor and store blood oxygen, heart rate, myoelectric, and EEG signals during cerebral palsy rehabilitation, and it can track and monitor signals during the rehabilitation treatment process. The system includes two parts: hardware design and software design. The hardware design includes a data signal acquisition module, a main control chip (ESP32), a muscle electrical sensor module, a brain electrical sensor module, a blood/heart rate acquisition module, etc. It is primarily for real-time signal data acquisition, processing, and uploading to the cloud server. The software design includes functions such as data receiving, data processing, data storage, network configuration, and remote communication and enables the visual monitoring of data signals. The system can achieve real-time monitoring of electromyography, electroencephalography, and blood oxygen levels, as well as the heart rate of patients with cerebral palsy, and adjust rehabilitation training in real-time during the rehabilitation process. At the same time, based on the real-time storage of the original electromyography and electroencephalography data, it can provide auxiliary guidance for later rehabilitation evaluation and effective data support for the entire rehabilitation treatment process. Full article

The cognitive state of a person can be categorized using the circumplex model of emotional states, a continuous model of two dimensions: arousal and valence. The purpose of this research is to select a machine learning model(s) to be integrated into a virtual reality (VR) system that runs cognitive remediation exercises for people with mental health disorders. As such, the prediction of emotional states is essential to customize treatments for those individuals. We exploit the Remote Collaborative and Affective Interactions (RECOLA) database to predict arousal and valence values using machine learning techniques. RECOLA includes audio, video, and physiological recordings of interactions between human participants. To allow learners to focus on the most relevant data, features are extracted from raw data. Such features can be predesigned, learned, or extracted implicitly using deep learners. Our previous work on video recordings focused on predesigned and learned visual features. In this paper, we extend our work onto deep visual features. Our deep visual features are extracted using the MobileNet-v2 convolutional neural network (CNN) that we previously trained on RECOLA’s video frames of full/half faces. As the final purpose of our work is to integrate our solution into a practical VR application using head-mounted displays, we experimented with half faces as a proof of concept. The extracted deep features were then used to predict arousal and valence values via optimizable ensemble regression. We also fused the extracted visual features with the predesigned visual features and predicted arousal and valence values using the combined feature set. In an attempt to enhance our prediction performance, we further fused the predictions of the optimizable ensemble model with the predictions of the MobileNet-v2 model. After decision fusion, we achieved a root mean squared error (RMSE) of 0.1140, a Pearson’s correlation coefficient (PCC) of 0.8000, and a concordance correlation coefficient (CCC) of 0.7868 on arousal predictions. We achieved an RMSE of 0.0790, a PCC of 0.7904, and a CCC of 0.7645 on valence predictions. Full article

Performance during training is a poor predictor of long-term retention. Worse yet, conditions of training that produce rapidly improving performance typically do not produce long-lasting, generalizable learning. As a result, learners and instructors alike can be misled into adopting training or educational experiences that are suboptimal for producing actual learning. Computer-based educational training platforms can counter this unfortunate tendency by providing only productive conditions of instruction—even if they are unintuitive (e.g., spacing instead of massing). The use of such platforms, however, introduces a different liability: being easy to interrupt. An assessment of this possible liability is needed given the enormous disruption to modern education brought about by COVID-19 and the subsequent widespread emergency adoption of computer-based remote instruction. The present study was therefore designed to (a) explore approaches for detecting interruptions that can be reasonably implemented by an instructor, (b) determine the frequency at which students are interrupted during a cognitive-science-based digital learning experience, and (c) establish the extent to which the pandemic and ensuing lockdowns affected students’ metacognitive ability to maintain engagement with their digital learning experiences. Outliers in time data were analyzed with increasing complexity and decreasing subjectivity to identify when learners were interrupted. Results indicated that only between 1.565% and 3.206% of online interactions show evidence of learner interruption. And although classroom learning was inarguably disrupted by the pandemic, learning in the present, evidence-based platform appeared to be immune. Full article

AI-driven mineral prospectivity mapping (MPM) is a valid and increasingly accepted tool for delineating the targets of mineral exploration, but it suffers from noisy and unrepresentative input features. In this study, a set of fractal and multifractal methods, including box-counting calculation, concentration–area fractal modeling, and multifractal analyses, were employed to excavate the underlying nonlinear mineralization-related information from geological features. Based on these methods, multiple feature selection criteria, namely prediction–area plot, K-means clustering, information gain, chi-square, and the Pearson correlation coefficient, were jointly applied to rank the relative importance of ore-related features and their fractal representations, so as to choose the optimal input feature dataset readily used for training predictive AI models. The results indicate that fault density, the multifractal spectrum width (∆α) of the Yanshanian intrusions, information dimension (D₁) of magnetic anomalies, correlation dimension (D₂) of iron-oxide alteration, and the D₂ of argillic alteration serve as the most effective predictor features representative of the corresponding ore-controlling elements. The comparative results of the model assessment suggest that all the AI models trained by the fractal datasets outperform their counterparts trained by raw datasets, demonstrating a significant improvement in the predictive capability of fractal-trained AI models in terms of both classification accuracy and predictive efficiency. A Shapley additive explanation was employed to trace the contributions of these features and to explain the modeling results, which imply that fractal representations provide more discriminative and definitive feature values that enhance the cognitive capability of AI models trained by these data, thereby improving their predictive performance, especially for those indirect predictor features that show subtle correlations with mineralization in the raw dataset. In addition, fractal-trained models can benefit practical mineral exploration by outputting low-risk exploration targets that achieve higher capturing efficiency and by providing new mineralization clues extracted from remote sensing data. This study demonstrates that the fractal representations of geological features filtered by multi-criteria feature selection can provide a feasible and promising means of improving the predictive capability of AI-driven MPM. Full article