Segmentation of medical images is a crucial step in many clinical applications, including the precise diagnosis and treatment of diseases like breast cancer. Therefore, automated segmentation of breast tumors from breast ultrasound images remains a challenging task. In this paper, we developed a new model, called Ultrasound Network (US-Net), which uses the U-Net architecture with attention gates embedded in the skip connections to assign weights to feature maps based on their importance for the segmentation task. Our method underwent evaluation on three public datasets: BUSI, UDIAT, and STUHospital, using the Dice coefficient as the primary metric for segmentation performance. Notably, US-Net achieved impressive Dice coefficients of 86.99%, 94.38%, and 94% on BUSI, UDIAT, and STUHospital, respectively. Experimental results showed that our network outperformed the latest image segmentation methods for lesion segmentation in breast ultrasound.

Simulation in healthcare originates from risky intervention. This training method is nowadays essential and aims to help medical students master medical and surgical gestures. Also, it reduces and limits medical error frequency. Cardiac catheterization allows doctors to place a stent inside a blocked artery in order to hold it open and prevent it from blocking again. This technique remains both frequent and delicate. Therefore, the use of a simulator before practicing it on real patients will allow doctors to acquire good gestural techniques. From this perspective, this paper deals with the creation of a stenting simulator called "Sim-Heart Abulcasis". Therefore, it is a two-part project, the first one consists of reconstructing a heart volume based on CT scans and then printing it in PLA material. The second is modeling a stent using SolidWorks software and then printing it in resin material. Finally, both printed pieces were fused to create a simulator prototype.

Simulation is widely used in medical schools, nursing programs, and other healthcare training institutions. This training method is nowadays essential and aims to reduce and limit the frequency of medical errors. Flexible ureteroscopy will allow the diagnosis and treatment of kidney stones located at the top of the urinary tract. This technique, which is efficient, reproducible, and not very traumatic, is very frequently used to treat nephrolithiasis often caused by poor hydration and/or nutrition. Complications of this surgical technique remain rare but the use of a simulator before the operation allows to initiate and prepare doctors in order to acquire good gestural techniques. From this perspective, this paper will deal with the creation and 3D printing of a urinary tract simulator named "UreteroSim-Abulcasis". For this purpose, uro-scans provided by the Cheikh Zaid International Hospital in Rabat (Morocco) were used to obtain a 3D reconstruction of the urinary tract. A list of modeling and 3D reconstruction simulators was first established to determine the one that would be the most suitable, and then a prototype was created using a 3D fused deposition modeling printer.

Melanoma the most dangerous type of skin cancer, has been on the rise in recent years. Hands-on identification of melanoma in its early stages with the unaided eye is error-prone and necessitates extensive expertise and experience. Due to the scarcity of skilled dermatologists, a computerized and automated technique is required to effectively identify melanoma. The following approach attempts to accomplish this task by creating a new approach capable of segmenting, then classifying melanoma. The procedure begins with the preparation of dermoscopic images to remove hairs using the Dull Razor algorithm, followed by image segmentation, in which we computed the Hausdorff Distance, Dice, and Jaccard coefficients to determine which channel of the RGB space was best to utilize to separate the skin lesion from the background. The segmented images using the green channel are then utilized to calculate the Gray Level Co-occurrence Matrices (GLCM) and to extract the color characteristics of the region of interest. Our approach is able to achieve a Dice coefficient and an accuracy of 95% on the PH2 dermoscopic images.

Skin cancer is a serious worldwide health worry with high mortality rates and high grimness. For this reason, to successfully diagnose skin lesions, a computer-aided automatic diagnostic system is required. One of the most crucial methods to do that is the segmentation of skin lesions. In this paper, we present a new model that integrates two architectures, the U-Net and the VGG19. Furthermore, to improve the results of segmentation, we also employ image preprocessing, including the Dull-Razor algorithm for hair removal and Contrast Limited Adaptive Histogram Equalization (CLAHE) to improve the image contrast. Moreover, we evaluated our model on three datasets: ISIC 2016, ISIC 2017, and ISIC 2018. Our suggested model achieved satisfactory results compared to the state-of-the-art.

Outlier and anomaly detection are widely used in several fields of study such as social networks, statistics, and knowledge discovery. In social networks, it is useful to detect structural abnormalities which are different from the typical behavior of the social network in order to maintain the network security and privacy. In this paper, we suggest a new approach for outlier and anomalous behavior detection in social networks based on combined graph pattern matching and constraint programming. For this purpose, we utilize the Constraint Programming (CP) techniques for matching the original graph data with the graph pattern data, to detect two formalized anomalies: anomalous nodes and anomalous edges. We also introduce a neighborhood constraint formalization that aims to precise the anomaly that replaced specific node as well as the changes that it made within the networks. Finally, we present our experimental results that show the effectiveness and efficiency of our approach in terms of computational time and matching accuracy.

3D object recognition and categorization are an important problem in computer vision field. Indeed, this is an area that allows many applications in diverse real problems as robotics, aerospace, automotive industry and food industry. Our contribution focuses on real 3D object recognition and categorization using the Deep Belief Networks method (DBN). We extract descriptors from cloud keypoints, then we train the resulting vectors with DBN. We evaluate the performance of this contribution on two datasets, Washington RGB-D object dataset and our own real 3D object dataset. The second one is built from real objects, following the same acquisition conditions than those used for Washington dataset acquisition. By this proposed approach, a DBN could be designed to treat the high-level features for real 3D object recognition and categorization. The experiment results on standard dataset show that our method outperforms the state-of-the-art used in the 3D object recognition and categorization.