Hmidi Alaeddine

Understanding how brain activity varies across neurological and neurodevelopmental disorders requires tools capable of revealing patterns hidden in complex electroencephalographic (EEG) data. Conditions such as epilepsy, Alzheimer’s disease, dementia, and autism exhibit distinct alterations in neural oscillations and connectivity, which remain difficult to interpret in real time; therefore, this study proposes an interactive interface for intuitive exploration and analysis of disease-specific EEG dynamics. The system integrates classical signal processing techniques and computational modeling to extract spectral features, inter-electrode coherence, and spatial activation patterns, which are visualized through spectrograms, topographic maps, and connectivity graphs that update continuously. In addition, a web-based platform is incorporated to enable clinicians and technicians to store and manage patient information, including diagnosis, severity level, number of recordings, sampling frequency, recording duration, and acquisition dates, supporting structured data organization and longitudinal monitoring. The results demonstrate that the interface captures meaningful differences between disorders, with epileptic patterns showing strong synchronization and burst activity, while neurodegenerative conditions exhibit spectral slowing and reduced connectivity. Overall, the proposed framework provides an effective and accessible tool for EEG visualization, combining interactive analysis with clinical data management to support research, education, and potential clinical applications.

Accurate classification of electrocardiogram (ECG) beats remains a major challenge due to noise, inter-patient variability, and spectral distortions that degrade signal quality, particularly in real-time monitoring and embedded system scenarios. To address these limitations, this paper proposes ECG-MSEE-Net, a novel deep learning model for ECG beat classification that integrates a convolutional neural network (CNN), an adaptive spectral weighting mechanism based on Maximum Spectral Energy Entropy (MSEE), and a Transformer encoder. The CNN extracts discriminative local morphological features, while the MSEE-based module adaptively weights ECG segments according to their spectral quality, enhancing robustness against noisy or atypical signals. Additionally, the Transformer encoder captures long-term temporal dependencies to further optimize classification performance. The proposed model is evaluated on the MIT-BIH Arrhythmia dataset using a patient-independent protocol. Experimental results demonstrate an accuracy of 98.74%, an F1-score of 92.19%, and strong generalization capabilities, even in the presence of noise. Comparisons with state-of-the-art methods confirm that ECG-MSEE-Net effectively handles spectral anomalies and achieves competitive performance. Furthermore, validation on an embedded platform (Raspberry Pi) highlights its potential for real-time ECG monitoring and connected healthcare applications.

Poultry farming plays a vital role in global food production, requiring efficient management to ensure productivity and animal welfare. Traditional methods, largely based on manual monitoring, are often inefficient, error-prone, and costly. With the rise of Internet of Things (IoT) technologies, intelligent systems now enable remote monitoring and management of environmental conditions, farm operations, and disease prevention. Platforms such as ThingSpeak allow for real-time data collection, processing, and visualization, offering a cost-effective solution for poultry farm management. By integrating sensors to measure temperature, humidity, air quality, and feeding, and by leveraging ThingSpeak’s analytical tools, farms can automatically adjust conditions and support proactive decision-making. This not only reduces operational costs but also improves efficiency, resource management, and animal health monitoring. The growing demand for poultry products has pressured farms to increase production, which heightens the risk of disease outbreaks and significant economic losses. Traditional disease detection methods, which depend on manual inspections by skilled professionals, are labor-intensive and delay timely intervention. To address these challenges, an IoT-based poultry disease detection and classification system is proposed. This system employs sensors for continuous health monitoring and artificial intelligence algorithms such as YOLOV7 and MobileNetV3 to analyze data. YOLOV7 segments regions of interest from automatically captured fecal images, while MobileNetV3 classifies them into four states: healthy, coccidiosis, salmonella, and Newcastle disease. Trained on Zenodo database samples, these models achieve high accuracy, providing farmers and veterinarians with an effective tool for proactive disease management and sustainable poultry farming.

Breast cancer remains one of the most significant health challenges, being the second leading cause of death among women worldwide. Early and accurate diagnosis is critical to improving treatment outcomes and increasing survival rates. In this study, we present an innovative application of the WRN-28-2 model, a deep convolutional neural network pre-trained on ImageNet, for the classification of histopathological breast cancer images from the BreakHis dataset. By leveraging transfer learning, the model was fine-tuned to differentiate between benign and malignant cases, achieving a remarkable classification accuracy of 99.16% on the test set. Moreover, the model outperformed existing state-of-the-art techniques on the same dataset. This research highlights the efficiency and adaptability of WRN-28-2 for medical image classification and opens avenues for its deployment in real-world diagnostic scenarios, particularly in resource-constrained environments.

This paper introduces a fine-tuned Convolutional Neural Network (CNN) for multiclass classification of brain tumors on contrast-enhanced T1-weighted MRI scans. The proposed model integrates batch normalization, dropout, and lightweight convolutional blocks to extract discriminative features while maintaining computational efficiency suitable for embedded deployment. Experiments were conducted on the Figshare dataset comprising 3,064 MRI slices from 233 patients with gliomas, meningiomas, or pituitary tumors. Images were preprocessed through resizing, normalization. The model was trained using the Adam optimizer with a learning rate of 1e-4, a batch size of 32, and 100 epochs. Evaluation metrics included accuracy, precision, recall, and F1-score. The fine-tuned CNN achieved an overall accuracy of 94.08%, with class-specific performance indicating strong results for pituitary tumors (precision 95.65%, recall 95.96%) and meningiomas (precision 90.20%, recall 88.81%), while glioma classification showed high sensitivity (recall 96.85%) but lower precision (75.00%). To validate real-world applicability, the model was converted to TensorFlow Lite and deployed on a Raspberry Pi 4, achieving an inference time of approximately 60 ms per image. These findings demonstrate that fine-tuned CNNs can offer a competitive and resource-efficient solution for computer-aided diagnosis of brain tumors, balancing accuracy and practicality in clinical environments with limited computational resources.