Adel Al-Jumaily

Electromyography (EMG)-based speech recognition provides a non-acoustic alternative in high-noise environments and serves as a prospective foundation for assistive technologies for users with speech impairments. While prior work has largely focused on model architecture, the impact of EMG feature representations on the robustness and subject-independent generalization of deep learning models remains underexplored. This study aims to investigate how feature representation influences subject-independent EMG speech recognition by comparing two feature extraction pipelines—a conventional Mel-Frequency Spectral Coefficients (MFSC) pipeline and a Wavelet Scattering Transform (WST)-based pipeline—across four temporal back ends: gated recurrent unit (GRU), Transformer, ReMamba, and BiMamba. Recordings from 100 participants during voiced (vocalized) speech reading a closed set of fixed sentences were used, of which 20 participants’ recordings were excluded from model development and used exclusively for subject-independent testing to verify generalization. The speaker dependence of both representations was additionally examined with a participant-level variance analysis. The WST-based pipeline achieved higher sentence-level accuracy than the MFSC pipeline with every backend, with the best result reaching 85.82% on unseen participants. For the Transformer backend, accuracy increased from 78.59% with MFSC to 85.82% with the WST pipeline. Speaker identity explained 12% of the feature variance in the WST pipeline compared with 69% for MFSC. When the filtering pipeline used for MFSC was applied before scattering extraction, accuracy decreased by 2.16 percentage points, indicating that this preprocessing may suppress EMG information useful to the model. Collectively, the findings identify the WST-based pipeline as a less speaker-dependent EMG feature representation that improves cross-subject performance across different sequence models and may provide a basis for intelligent EMG speech-recognition systems.

Early and accurate diagnosis of brain diseases such as dementia, autism, and right hemisphere damage is crucial for effective intervention and treatment. Traditional diagnostic approaches often rely on clinical assessments, which can be time-consuming and subjective. In this study, we propose a methodology leveraging graph neural networks (GNNs) for brain disease classification based on speech transcripts or textual data. Our approach incorporates a subject similarity graph with an entropy-guided reinforcement regularisation mechanism to model complex linguistic patterns and capture inter-subject relationships, improving discrimination between neurological conditions. Experimental results demonstrate consistent performance gains over strong baseline models. The proposed framework provides a cost-effective and computationally efficient diagnostic tool with potential applicability to early-stage clinical decision support.

Following major accidents or illness, some patients may lose partially or completely some of their mental or physical function. Rehabilitation approaches and systems play an important role in patients' recovery. This manuscript focuses on rehabilitation systems. We provide insights on major and recent approaches, signals, and systems used to assist patient recovery. In order to help researchers, engineers, or physicians, this paper provides a comprehensive survey on recent approaches and systems used in the rehabilitation process. By highlighting the advantages and drawbacks of existing rehabilitation systems, this contribution can support initial studies or projects and encourage young researchers, as well as biomedical companies, to investigate and invest in new technologies. We hope that our study will participate in the global efforts paid to reduce patient suffering and help health care centers in their activities. To achieve this goal, we consider in our study recent technologies such as machine learning, artificial intelligence, signal processing, without missing to highlight various biomedical signals and exploring several developed systems.

Social robots have the potential to revolutionize the way we interact with technology, providing a wide range of services and applications in various domains, such as healthcare, education, and entertainment. However, most existing social robotics platforms are operated based on embedded computers, which limits the robot’s capabilities to access advanced AI-based platforms available online and which are required for sophisticated physical human–robot interactions (such as Google Cloud AI, Microsoft Azure Machine Learning, IBM Watson, ChatGPT, etc.). In this research project, we introduce a cloud-based framework that utilizes the benefits of cloud computing and clustering to enhance the capabilities of social robots and overcome the limitations of current embedded platforms. The proposed framework was tested in different robots to assess the general feasibility of the solution, including a customized robot, “BuSaif”, and commercialized robots, “Husky”, “NAO”, and “Pepper”. Our findings suggest that the implementation of the proposed platform will result in more intelligent and autonomous social robots that can be utilized by a broader range of users, including those with less expertise. The present study introduces a novel methodology for augmenting the functionality of social robots, concurrently simplifying their utilization for non-experts. This approach has the potential to open up novel possibilities within the domain of social robotics.

The infectious nature of the COVID-19 virus demands rapid detection to quarantine the infected to isolate the spread or provide the necessary treatment if required. Analysis of COVID-19-infected chest Computed Tomography Scans (CT scans) have been shown to be successful in detecting the disease, making them essential in radiology assessment and screening of infected patients. Single-model Deep CNN models have been used to extract complex information pertaining to the CT scan images, allowing for in-depth analysis and thereby aiding in the diagnosis of the infection by automatically classifying the chest CT scan images as infected or non-infected. The feature maps obtained from the final convolution layer of the Deep CNN models contain complex and positional encoding of the images’ features. The ensemble modeling of these Deep CNN models has been proved to improve the classification performance, when compared to a single model, by lowering the generalization error, as the ensemble can meta-learn from a broader set of independent features. This paper presents Deep Ensemble Learning models to synergize Deep CNN models by combining these feature maps to create deep feature vectors or deep feature maps that are then trained on meta shallow and deep learners to improve the classification. This paper also proposes a novel Attentive Ensemble Model that utilizes an attention mechanism to focus on significant feature embeddings while learning the Ensemble feature vector. The proposed Attentive Ensemble model provided better generalization, outperforming Deep CNN models and conventional Ensemble learning techniques, as well as Shallow and Deep meta-learning Ensemble CNNs models. Radiologists can use the presented automatic Ensemble classification models to assist identify infected chest CT scans and save lives.

Social distancing is one of the most important ways to prevent many diseases, especially the respiratory system, where the latest internationally spread is coronavirus disease, and it will not be the last. The spreading of this pandemic has become a major threat to human life, especially to the elderly and people suffering from chronic diseases. During the Corona pandemic, medical authorities were keen to control the spread through social distancing and monitoring it in markets, universities, and schools. This monitoring was mostly used to estimate the distance with the naked eye and interfere with estimating the distance on the observer only. In this study, a computer application was designed to monitor social distancing in closed areas, especially in schools and kindergartens, using a fast, effective and unobtrusive technique for children. In addition to this system, we use augmented reality to help to determine the location of violation of social distancing. This system was tested, and the results were accurate exceeding 98.5%.

Bio-signals pattern recognition systems can be impacted by several factors with a potential to limit their associated performance and clinical translation. Among these factors, selecting the optimum feature extraction method, that can effectively exploit the interaction between the temporal and spatial information, is the most prominent. Despite the potential of deep learning (DL) models for extracting temporal, spatial, or temporal-spatial information, they are typically restricted by their need for a large amount of training data. The deep wavelet scattering transform (WST) is a relatively recent advancement within the DL literature to replace expensive convolution neural networks models with computationally less demanding methods. However, while some studies have used WST to extract features from biological signals, it has not been investigated before for electromyogram (EMG) and electroencephalogram (EEG) signals feature extraction. To investigate the hypothesis of the usefulness of WST for processing EMG and EEG signals, this study used a tremor dataset collected by the authors from people with tremor disorders. Specifically, the proposed work achieved three goals: (a) study the performance of extracting features from low-density EMG signals (8 channels), using the WST approach, (b) study the effect of extracting the features from high-density EEG signals (33 channels), using WST and study its robustness against changing the spatial and temporal aspects of classification accuracy, and (c) classify tremor severity using the WST method and compare the results with other well-known feature extraction approaches. The classification error rates were significantly reduced (maximum of nearly 12%) compared with other feature sets.