| Titre : | A Deep learning approach for children pneumonia early diagnosis |
| Auteurs : | Remaigui REMAIGUI, Auteur ; Laïd Kahloul, Directeur de thèse |
| Type de document : | Monographie imprimée |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2022 |
| Format : | 1 vol. (55 p.) / couv. ill. en coul |
| Langues: | Anglais |
| Mots-clés: | Deep learning, neural convolutional networks, early diagnosis, pneumonia, smart healthcare |
| Résumé : |
Pneumonia is a bacterial or viral illness that causes inflammation in the lungs, it strongly affects children. Early diagnosis is an important factor in terms of the successful treatment process.
Generally, a professional radiologist can diagnose the condition by checking out the chest X-ray pictures. Nevertheless, it might be ambiguous or mistaken with other conditions as well as it can make the diagnosis subjective. Therefore, computer-aided diagnosis systems are needed to guide the clinicians. In this master project, we have utilized the convolutional neural network (CNN) technique to build a model that can detect the pneumonia for children. We used this model because CNNbased deep learning classification algorithms can automatically extract high-level representations. The model was trained, validated, and tested using a public data set, and the obtained results are considered satisfactory since they demonstrated an improvement compared to the existing works which have used the same data set. . |
| Sommaire : |
Acknowledgements i Abstract ii Résumé iii List of Figures x List of Tables 0 1 General Introduction 1 2 Background of the work 3 2.1 Introduction 3 2.2 Pneumonia 3 2.2.1 The lungs 3 2.2.2 Pneumonia disease 6 2.3 presentation of X-ray images 7 2.3.1 What is a chest X-ray? 7 2.3.2 Functioning of X-ray images 7 2.3.3 Uses of Chest x-rays 7 2.4 Machine learning 8 2.4.1 Machine learning methods 9 2.5 Deep learning 12 2.5.1 Artificial neural networking 13 2.5.2 Convolutional Neural Network 17 2.5.3 Popular CNN Architectures 20 2.5.4 Transfer learning . 23 2.6 Healthcare 23 2.6.1 Definition . 23 2.6.2 Stages 23 2.6.3 Artificial intelligence and healthcare 25 2.7 Related works 26 2.8 Conclusion 28 3 Design of a deep learning architecture for pneumonia detection in pediatric 29 3.1 Introduction 29 3.2 System design 29 3.2.1 Data preparation 30 3.2.2 Training Model . 32 3.2.3 Testing Model . 34 3.3 Conclusion 35 4 Implementation and results 36 4.1 Introduction 36 4.2 Development environments and tools 36 4.2.1 Google colab 36 4.2.2 Python 36 4.2.3 TensorFlow 37 4.2.4 Keras 37 4.2.5 Numpy 37 4.2.6 Matplotlib 38 4.2.7 Kaggle 38 4.3 Developing the back-end 38 4.3.1 Dataset preparation 38 4.3.2 Building our CNN Model 39 4.3.3 Testing our CNN Model 46 4.4 Empirical evaluation and results 4.4.1 Results of the training phase 48 4.4.2 Model evaluation on test data 49 4.4.3 Testing with the first dataset. 49 4.4.4 Test for second dataset 50 4.4.5 Table of Comparison 51 4.5 Conclusion 52 5 Conclusion and Perspectives 53 5.1 Conclusion . 53 5.2 Perspectives 53 References 55 |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| MINF/729 | Mémoire master | bibliothèque sciences exactes | Consultable |
