Browsing by Author "Mahmoudi, Said"

Filter results by typing the first few letters
Now showing 1 - 4 of 4
  • No Thumbnail Available
    Item
    GAN data augmentation for improved automated atherosclerosis screening from coronary CT angiography
    (2023) Laidi, Amel; Ammar, Mohammed; El Habib Daho, Mostafa; Mahmoudi, Said
    Atherosclerosis is a chronic medical condition that can result in coronary artery disease,strokes, or even heart attacks. early detection can result in timely interventions and save lives.OBJECTIVES: In this work, a fully automatic transfer learning-based model was proposed for Atherosclerosisdetection in coronary CT angiography (CCTA). The model’s performance was improved by generating trainingdata using a Generative Adversarial Network.METHODS: A first experiment was established on the original dataset with a Resnet network, reaching 95.2%accuracy, 60.8% sensitivity, 99.25% specificity and 90.48% PPV. A Generative Adversarial Network (GAN) wasthen used to generate a new set of images to balance the dataset, creating more positive images. Experimentswere made adding from 100 to 1000 images to the dataset.RESULTS: adding 1000 images resulted in a small drop in accuracy to 93.2%, but an improvement in overallperformance with 89.0% sensitivity, 97.37% specificity and 97.13% PPV.CONCLUSION: This paper was one of the early research projects investigating the efficiency of dataaugmentation using GANs for atherosclerosis, with results comparable to the state of the art
  • No Thumbnail Available
    Item
    An improved system for emphysema recognition using CNN features extraction and AdaBoost-Decision treeclassifier
    (2021) Ammar, Mohammed; Mahmoudi, Said
    In this work, a hybrid model composed of a CNN and a classical machine learning methodwas proposed to improve the classification of emphysema diseases. Firstly, we have proposeda pre-treatment step based on contrast adjustment in order to improve the performancesof the proposed model. Second, we extract the features from the deeper layers of the CNNclassifier, then we classify these features with decision tree and AdaBoost algorithm. Theproposed model is validated by usinga set of 168 manually annotated ROIs for each CTimage, comprising the three classes: normal tissue, centrilobular emphysema, and para-septal emphysema. The obtained results show that the hybrid model proposed in thiswork provides the best accuracy in the case of the AdaBoost-Decision Tree classifier.Acomparison with CNN, CNN-SVM and CNN-AdaBoost-Decision Tree classifier has beenperformed. As conclusion, the CNN-AdaBoost-Decision Tree classifier provide the bestresults with an accuracy of 100%
  • No Thumbnail Available
    Item
    Smart embedded system for sleep apnea monitoring from ECG signals
    (American Institute of Physics, 2023) Ammar, Mohammed; Messaoudi, Noureddine; Faked, Djouher; Noui, Rima; Mahmoudi, Said
    In this paper, an intelligent monitoring system was proposed to follow vital parameters such as the electrocardiogram (ECG), oxygen saturation (SPO2), the temperature of the patient, and also heart rate. The system is built around a Raspberry 3B+ and an Arduino Uno. The prototype is equipped with an intelligent system that can currently detect sleep apnea from ECG signals. These parameters are detected by the following sensors: AD8232, and MAX 30102. We have implemented and compared three algorithms: Perceproron multi-layer, Support Vector Machine, and a Random Forest Classifier.
  • No Thumbnail Available
    Item
    Toward an automatic detection of cardiac structures in short and long axis views
    (Elsevier, 2023) Laidi, Amel; Mohammed, Ammar; El Habib Daho, Mostafa; Mahmoudi, Said
    Objective: This work aims to create an automatic detection process of cardiac structures in both short-axis and long-axis views. A workflow inspired by human thinking process, for better explainability. Methods: we began by separating the images into two classes: long axis and short axis, using a Residual Network model. Then, we used Particle Swarm Optimization for general segmentation. After segmentation, a characterization step based on shape descriptors calculated from bounding box and ANOVA for features selection were applied on the binary images to detect the location of each region of interest: lung, left and right ventricle in the short-axis view, the aorta, the left heart (left atrium and ventricle), and the right heart (right atrium and ventricle) in the long axis view. Results: we achieved a 90% accuracy on view separation. We have selected: Elongation, Compactness, Circularity, Type Factor, for short axis identification; and:Area, Centre of Mass Y, Moment of Inertia XY, Moment of Inertia YY, for long axis identification. Conclusion: a successful separation of long axis and short axis views allows for a better characterization and detection of segmented cardiac structures. After that, any method can be applied for segmentation, attribute selection, and classification. Significance: an attempt to introduce explainability into cardiac image segmentation, we tried to mimic the human workflow while computerizing each step. The process seems to be valid and added clarity and interpretability to the detection