Expert-guided extraction of relevant informations: Application to medical pathology detection / Hafida Chellakh

Public ISBD Titre : Expert-guided extraction of relevant informations: Application to medical pathology detection Type de document : document électronique Auteurs : Hafida Chellakh, Auteur ; Abdelouahab Moussaoui, Directeur de thèse Editeur : Sétif:UFA1 Année de publication : 2025 Importance : 1 vol (120 f.) Format : 29 cm Langues : Anglais (eng) Catégories : Thèses & Mémoires:Informatique Mots-clés : MRI Brain tumor Deep Learning Feature extraction BSIF descriptor DRB Classifier Index. décimale : 004 - Informatique Résumé : Magnetic Resonance Imaging (MRI) brain tumor identification and classification are costly and time-consuming due to tumor complexity and reliance on radiologist expertise. To overcome these challenges, automating the process is essential. This thesis leverages the power of deep learning for brain tumor analysis, presenting two key contributions. In the first contribution, we introduce an efficient model titled "Deep Rule-Based Classifier using Bank of Binarized Statistical Image Features (DRB-BBSIF)". This approach addresses the limitations of conventional MRI brain tumor diagnosis by offering a model that improves classification performance while reducing the complexity of the diagnostic process. The model explores the BSIF image descriptor for the feature extraction phase, Furthermore, to enhance its performance, we have constructed a Bank-BSIF, which is founded by the best parameters of BSIF filters. For the classification phase, we employed a deep rule-based (DRB) classifier. The DRB classifier functions through a self-organized set of IF-THEN fuzzy rules, guided by prototypes. These fuzzy rules, generated by the DRB classifier, serve as the classifier's core decision-making mechanism. The second contribution titled “MRI Brain Tumor Identification and Classification using Deep Learning Techniques” focuses on the synergistic integration of deep learning and rulebased classification. We propose a novel, simple, and automatic DRB-based scheme for MRI brain tumor classification. This model leverages the power of deep learning for feature extraction and combines it with the effectiveness of DRB for classification. The framework consists of three stages: preprocessing, feature extraction, and classification. Feature extraction utilizes deep learning networks like AlexNet, VGG-16, ResNet-50, and ResNet- 18 to extract features from the MRI images. A DRB classifier then utilizes these deep features for classification. Both methods are evaluated on publicly available datasets and demonstrate significant performance in classifying brain tumors (presence or absence) and even tumor types (multiclass). They outperform traditional techniques, highlighting their effectiveness in MRI brain tumor analysis. The thesis provides significant advancements in MRI brain tumor identification and classification using deep learning techniques, presenting promising tools for computer-aided diagnosis. It also contributes to enhancing early disease detection and improving the efficiency and outcomes of treatment. Note de contenu : Sommaire GENERAL INTRODUCTION ............................................................................................................. 1 1. CONTEXT AND MOTIVATION .................................................................................................... 1 2. PROBLEM STATEMENT............................................................................................................ 1 3. THESIS OBJECTIVES ................................................................................................................ 3 4. THESIS CONTRIBUTIONS .......................................................................................................... 3 5. THESIS ORGANIZATION ........................................................................................................... 5 I CHAPTER I MRI FOR BRAIN TUMOR DIAGNOSIS .................................................................. 6 I.1 INTRODUCTION ..................................................................................................................... 6 I.2 BRAIN TUMORS .................................................................................................................... 6 I.2.1 WHAT IS BRAIN TUMOR? ............................................................................................... 6 I.2.2 TYPES OF BRAIN TUMORS ............................................................................................... 6 I.2.3 IMPACT OF BRAIN TUMOR ON THE HEALTH ......................................................................... 8 I.2.3.1 Benign Tumors ..................................................................................... 8 I.2.3.2 Malignant Tumors ................................................................................ 8 I.2.3.3 Metastatic Tumors ............................................................................... 9 I.3 MAGNETIC RESONANCE IMAGING – MRI – AN OVERVIEW .......................................................... 10 I.3.1 INTRODUCTION.......................................................................................................... 10 I.3.2 OVERVIEW OF VARIOUS IMAGING METHODS .................................................................... 10 I.3.2.1 X-ray .................................................................................................. 11 I.3.2.2 Tomographic Imaging......................................................................... 11 I.3.3 MAGNETIC RESONANCE IMAGING (MRI) ........................................................................ 13 I.3.3.1 Advantages of MRI ............................................................................. 14 I.3.3.2 Disadvantages of MRI......................................................................... 14 I.3.3.3 Basic Principles of MRI ....................................................................... 15 I.4 CONCLUSION ...................................................................................................................... 18 II. CHAPTER II MATERIALS AND METHODS .............................................................................. 19 II.1 INTRODUCTION ................................................................................................................... 19 II.2 ARTIFICIAL INTELLIGENCE, MACHINE LEARNING AND DEEP LEARNING PARADIGM .............................. 19 II.3 MACHINE LEARNING (ML): ................................................................................................... 20 II.3.1 DEFINITION .............................................................................................................. 20 II.3.2 TYPES OF MACHINE LEARNING ...................................................................................... 20 II.3.2.1 Supervised Learning: .......................................................................... 20 II.3.2.2 Unsupervised Learning ....................................................................... 20 II.3.2.3 Reinforcement Learning ..................................................................... 21 II.3.3 CONCEPTS OF MACHINE LEARNING MODELS .................................................................... 21 II.4 DEEP LEARNING (DL) ........................................................................................................... 22 II.4.1 DEFINITION .............................................................................................................. 22 II.4.2 HISTORY OF DEEP LEARNING ........................................................................................ 22 II.4.3 CONCEPTS OF DEEP LEARNING MODELS .......................................................................... 24 II.4.3.1 Neurons ............................................................................................. 24 II.4.3.2 Activation Functions: The Decision Maker .......................................... 25 II.4.3.3 Artificial Neural Network .................................................................... 29 II.4.4 TYPES OF ARTIFICIAL NEURAL NETWORK .......................................................................... 30 II.4.4.1 Perceptron ......................................................................................... 30 II.4.4.2 Multilayer Perceptron (MLP) .............................................................. 30 II.4.4.3 Convolutional Neural Network (CNN) ................................................. 30 II.4.4.4 Recurrent Neural Network (RNN) ....................................................... 31 II.4.4.5 Long Short-Term Memory (LSTM) ...................................................... 32 II.4.4.6 Generative Adversarial Networks (GANs) ........................................... 32 II.4.4.7 Sequence to Sequence Models (Seq2Seq) .......................................... 33 II.5 DEEP LEARNING IN MEDICAL IMAGE ANALYSIS: .......................................................................... 34 II.5.1 THE IMPACT OF DEEP LEARNING ON MEDICAL IMAGE ANALYSIS ........................................... 34 II.5.2 APPLICATIONS IN MEDICAL IMAGING.............................................................................. 34 II.5.3 DL TECHNIQUES FOR MEDICAL IMAGE ANALYSIS .............................................................. 35 II.5.4 ADVANTAGES OF DEEP LEARNING IN MEDICAL IMAGING .................................................... 37 II.5.5 CHALLENGES AND LIMITATIONS ..................................................................................... 37 II.5.6 FUTURE DIRECTIONS ................................................................................................... 37 II.5.7 CNNS FOR MRI BRAIN TUMOR CLASSIFICATION ............................................................... 38 II.5.7.1 Input layer ......................................................................................... 39 II.5.7.2 Convolutional Layer ........................................................................... 39 II.5.7.3 Pooling Layer ..................................................................................... 40 II.5.7.4 Fully Connected Layer ........................................................................ 41 II.5.7.5 Logistic or Softmax Layer (LOSS) ......................................................... 41 II.5.7.6 Output Layer: The Final Decision Maker ............................................. 41 II.6 PERFORMANCE EVALUATION MEASUREMENTS ......................................................................... 42 II.6.1 CONFUSION MATRIX................................................................................................... 42 II.6.2 ACCURACY................................................................................................................ 42 II.6.3 SENSITIVITY .............................................................................................................. 42 II.6.4 SPECIFICITY ............................................................................................................... 43 II.6.5 F1-SCORE................................................................................................................. 43 II.6.6 ROC CURVE .............................................................................................................. 43 II.7 CONCLUSION ...................................................................................................................... 44 III. CHAPTER III BRAIN TUMOR CLASSIFICATION ...................................................................... 45 III.1 INTRODUCTION ................................................................................................................... 45 III.2 TAXONOMY OF MRI BRAIN IMAGE CLASSIFICATION ALGORITHMS ................................................ 45 III.3 CLASSIFICATION PROCESS ..................................................................................................... 48 III.3.1 PRE-PROCESSING STEP ................................................................................................ 49 III.3.2 FEATURE EXTRACTION STEP .......................................................................................... 49 III.3.2.1 Binarized Statistical Image Features (BSIF) ......................................... 50 III.3.2.2 Histogram of Oriented Gradients (HOG) ............................................. 51 III.3.2.3 GIST Descriptor: A Low-Dimensional Image Representation ............... 51 III.3.2.4 AlexNet .............................................................................................. 52 III.3.2.5 VGG-16 .............................................................................................. 52 III.3.2.6 Residual Network (ResNet-50 and ResNet-18) ................................ 53 III.3.3 CLASSIFICATION STEP: ................................................................................................. 53 III.3.3.1 Naive Bayes ....................................................................................... 54 III.3.3.2 K-Nearest Neighbor............................................................................ 54 III.3.3.3 Support Vector Machine (SVM) .......................................................... 55 III.3.3.4 Decision Trees .................................................................................... 57 III.4 OVERVIEW OF BRAIN TUMOR CLASSIFICATION .......................................................................... 58 III.4.1 MACHINE LEARNING TECHNIQUES ................................................................................. 58 III.4.2 DEEP LEARNING TECHNIQUES ....................................................................................... 59 III.5 DEEP RULE BASED CLASSIFIER FOR MRI BRAIN TUMOR CLASSIFICATION ........................................ 60 III.5.1 GENERAL ARCHITECTURE OF THE DRB CLASSIFIER ............................................................. 60 III.5.2 MASSIVELY PARALLEL FRB ........................................................................................... 62 III.5.2.1 Training process of the DRB system.................................................... 63 III.5.2.2 Validation process of the DRB system ................................................ 65 III.5.2.3 Decision Maker .................................................................................. 65 III.6 CONCLUSION ...................................................................................................................... 67 IV. CHAPTER IV: DRB-BBSIF FOR BRAIN TUMOR CLASSIFICATION ............................................. 68 IV.1 INTRODUCTION ................................................................................................................... 68 IV.2 THE ARCHITECTURE OF PROPOSED DRB-BBSIF CLASSIFIER ......................................................... 68 IV.2.1 EXTRACTION OF THE REGION OF INTEREST (ROI) .............................................................. 69 IV.2.2 EXPLORING BINARIZED STATISTICAL IMAGE FEATURES (BSIF) .............................................. 69 IV.2.3 DEEP RULE-BASED CLASSIFIER FOR MRI BRAIN TUMOR CLASSIFICATION ................................ 71 IV.3 EXPERIMENTS AND RESULTS .................................................................................................. 72 IV.3.1 DATABASE ................................................................................................................ 72 IV.3.2 EXPERIMENT 1 – CONSTRUCTION OF BANK OF BSIF FILTERS ............................................... 73 IV.3.2.1 Objectives and Methodology ............................................................. 73 IV.3.2.2 Analysis of Results.............................................................................. 74 IV.3.2.3 Interpretation of Results .................................................................... 76 IV.3.2.4 Key findings ....................................................................................... 78 IV.3.3 EXPERIMENT 2: IMPACT OF FEATURE EXTRACTOR METHODS ............................................... 78 IV.3.3.1 Objective of Experiment 2 .................................................................. 78 IV.3.3.2 Analysis of results .............................................................................. 78 IV.3.3.3 Interpretation of results ..................................................................... 79 IV.3.3.4 Key findings ....................................................................................... 80 IV.3.4 EXPERIMENT 3: EVALUATION OF THE DRB-BBSIF MODEL .................................................. 80 IV.3.4.1 Objective of Experiment 3 .................................................................. 80 IV.3.4.2 Analysis of Results.............................................................................. 80 IV.3.4.3 Interpretation of Results .................................................................... 82 IV.3.4.4 Key Finding ........................................................................................ 83 IV.4 CONCLUSION ...................................................................................................................... 83 V. CHAPTER V: DRB WITH DEEP FEATURE EXTRACTION ........................................................... 85 V.1 INTRODUCTION ................................................................................................................... 85 V.2 PROPOSED METHODOLOGY ................................................................................................... 85 V.2.1 PRE-PROCESSING STEP ................................................................................................ 86 V.2.2 FEATURE EXTRACTION STEP .......................................................................................... 86 V.2.3 CLASSIFICATION STEP .................................................................................................. 86 V.3 DATABASE ......................................................................................................................... 86 V.4 EXPERIMENTS AND RESULTS .................................................................................................. 87 V.4.1 EXPERIMENT 1: ALEXNET WITH 4 DIFFERENT CLASSIFIERS ................................................... 88 V.4.1.1 Analysis of Results.............................................................................. 89 V.4.1.2 Key Finding ........................................................................................ 91 V.4.2 EXPERIMENT 2: VGG-16 WITH 4 DIFFERENT CLASSIFIERS.................................................... 92 V.4.2.1 Analysis of results .............................................................................. 93 V.4.2.2 Key Findings ....................................................................................... 94 V.4.3 EXPERIMENT 3: RESNET-50 WITH 4 DIFFERENT CLASSIFIERS ................................................ 96 V.4.3.1 Analysis of Results.............................................................................. 97 V.4.3.2 Key Findings ....................................................................................... 98 V.4.4 EXPERIMENT 4: RESNET-18 WITH 4 DIFFERENT CLASSIFIERS .............................................. 100 V.4.4.1 Analysis of Results............................................................................ 100 V.4.4.2 Key findings ..................................................................................... 103 V.4.5 COMPREHENSIVE ANALYSIS OF RESULTS ........................................................................ 103 V.4.5.1 Classifiers: Strengths and Challenges ................................................ 104 V.4.5.2 Deep Learning Features ................................................................... 105 V.5 COMPARISON BETWEEN THE TWO CONTRIBUTIONS .................................................................. 108 V.6 HIGH PERFORMANCE OF THE DRB CLASSIFIER......................................................................... 109 V.6.1 CHARACTERISTICS OF DRB CLASSIFIER .......................................................................... 109 V.6.2 PERFORMANCE METRICS OF THE DRB CLASSIFIER ........................................................... 110 V.6.3 COMPARISON WITH OTHER CLASSIFIERS........................................................................ 110 V.7 CONCLUSION .................................................................................................................... 111 CONCLUSION AND PERSPECTIVES ............................................................................................. 112 Côte titre : DI/0090 Expert-guided extraction of relevant informations: Application to medical pathology detection [document électronique] / Hafida Chellakh, Auteur ; Abdelouahab Moussaoui, Directeur de thèse . - [S.l.] : Sétif:UFA1, 2025 . - 1 vol (120 f.) ; 29 cm. Langues : Anglais (eng) Catégories : Thèses & Mémoires:Informatique Mots-clés : MRI Brain tumor Deep Learning Feature extraction BSIF descriptor DRB Classifier Index. décimale : 004 - Informatique Résumé : Magnetic Resonance Imaging (MRI) brain tumor identification and classification are costly and time-consuming due to tumor complexity and reliance on radiologist expertise. To overcome these challenges, automating the process is essential. This thesis leverages the power of deep learning for brain tumor analysis, presenting two key contributions. In the first contribution, we introduce an efficient model titled "Deep Rule-Based Classifier using Bank of Binarized Statistical Image Features (DRB-BBSIF)". This approach addresses the limitations of conventional MRI brain tumor diagnosis by offering a model that improves classification performance while reducing the complexity of the diagnostic process. The model explores the BSIF image descriptor for the feature extraction phase, Furthermore, to enhance its performance, we have constructed a Bank-BSIF, which is founded by the best parameters of BSIF filters. For the classification phase, we employed a deep rule-based (DRB) classifier. The DRB classifier functions through a self-organized set of IF-THEN fuzzy rules, guided by prototypes. These fuzzy rules, generated by the DRB classifier, serve as the classifier's core decision-making mechanism. The second contribution titled “MRI Brain Tumor Identification and Classification using Deep Learning Techniques” focuses on the synergistic integration of deep learning and rulebased classification. We propose a novel, simple, and automatic DRB-based scheme for MRI brain tumor classification. This model leverages the power of deep learning for feature extraction and combines it with the effectiveness of DRB for classification. The framework consists of three stages: preprocessing, feature extraction, and classification. Feature extraction utilizes deep learning networks like AlexNet, VGG-16, ResNet-50, and ResNet- 18 to extract features from the MRI images. A DRB classifier then utilizes these deep features for classification. Both methods are evaluated on publicly available datasets and demonstrate significant performance in classifying brain tumors (presence or absence) and even tumor types (multiclass). They outperform traditional techniques, highlighting their effectiveness in MRI brain tumor analysis. The thesis provides significant advancements in MRI brain tumor identification and classification using deep learning techniques, presenting promising tools for computer-aided diagnosis. It also contributes to enhancing early disease detection and improving the efficiency and outcomes of treatment. Note de contenu : Sommaire GENERAL INTRODUCTION ............................................................................................................. 1 1. CONTEXT AND MOTIVATION .................................................................................................... 1 2. PROBLEM STATEMENT............................................................................................................ 1 3. THESIS OBJECTIVES ................................................................................................................ 3 4. THESIS CONTRIBUTIONS .......................................................................................................... 3 5. THESIS ORGANIZATION ........................................................................................................... 5 I CHAPTER I MRI FOR BRAIN TUMOR DIAGNOSIS .................................................................. 6 I.1 INTRODUCTION ..................................................................................................................... 6 I.2 BRAIN TUMORS .................................................................................................................... 6 I.2.1 WHAT IS BRAIN TUMOR? ............................................................................................... 6 I.2.2 TYPES OF BRAIN TUMORS ............................................................................................... 6 I.2.3 IMPACT OF BRAIN TUMOR ON THE HEALTH ......................................................................... 8 I.2.3.1 Benign Tumors ..................................................................................... 8 I.2.3.2 Malignant Tumors ................................................................................ 8 I.2.3.3 Metastatic Tumors ............................................................................... 9 I.3 MAGNETIC RESONANCE IMAGING – MRI – AN OVERVIEW .......................................................... 10 I.3.1 INTRODUCTION.......................................................................................................... 10 I.3.2 OVERVIEW OF VARIOUS IMAGING METHODS .................................................................... 10 I.3.2.1 X-ray .................................................................................................. 11 I.3.2.2 Tomographic Imaging......................................................................... 11 I.3.3 MAGNETIC RESONANCE IMAGING (MRI) ........................................................................ 13 I.3.3.1 Advantages of MRI ............................................................................. 14 I.3.3.2 Disadvantages of MRI......................................................................... 14 I.3.3.3 Basic Principles of MRI ....................................................................... 15 I.4 CONCLUSION ...................................................................................................................... 18 II. CHAPTER II MATERIALS AND METHODS .............................................................................. 19 II.1 INTRODUCTION ................................................................................................................... 19 II.2 ARTIFICIAL INTELLIGENCE, MACHINE LEARNING AND DEEP LEARNING PARADIGM .............................. 19 II.3 MACHINE LEARNING (ML): ................................................................................................... 20 II.3.1 DEFINITION .............................................................................................................. 20 II.3.2 TYPES OF MACHINE LEARNING ...................................................................................... 20 II.3.2.1 Supervised Learning: .......................................................................... 20 II.3.2.2 Unsupervised Learning ....................................................................... 20 II.3.2.3 Reinforcement Learning ..................................................................... 21 II.3.3 CONCEPTS OF MACHINE LEARNING MODELS .................................................................... 21 II.4 DEEP LEARNING (DL) ........................................................................................................... 22 II.4.1 DEFINITION .............................................................................................................. 22 II.4.2 HISTORY OF DEEP LEARNING ........................................................................................ 22 II.4.3 CONCEPTS OF DEEP LEARNING MODELS .......................................................................... 24 II.4.3.1 Neurons ............................................................................................. 24 II.4.3.2 Activation Functions: The Decision Maker .......................................... 25 II.4.3.3 Artificial Neural Network .................................................................... 29 II.4.4 TYPES OF ARTIFICIAL NEURAL NETWORK .......................................................................... 30 II.4.4.1 Perceptron ......................................................................................... 30 II.4.4.2 Multilayer Perceptron (MLP) .............................................................. 30 II.4.4.3 Convolutional Neural Network (CNN) ................................................. 30 II.4.4.4 Recurrent Neural Network (RNN) ....................................................... 31 II.4.4.5 Long Short-Term Memory (LSTM) ...................................................... 32 II.4.4.6 Generative Adversarial Networks (GANs) ........................................... 32 II.4.4.7 Sequence to Sequence Models (Seq2Seq) .......................................... 33 II.5 DEEP LEARNING IN MEDICAL IMAGE ANALYSIS: .......................................................................... 34 II.5.1 THE IMPACT OF DEEP LEARNING ON MEDICAL IMAGE ANALYSIS ........................................... 34 II.5.2 APPLICATIONS IN MEDICAL IMAGING.............................................................................. 34 II.5.3 DL TECHNIQUES FOR MEDICAL IMAGE ANALYSIS .............................................................. 35 II.5.4 ADVANTAGES OF DEEP LEARNING IN MEDICAL IMAGING .................................................... 37 II.5.5 CHALLENGES AND LIMITATIONS ..................................................................................... 37 II.5.6 FUTURE DIRECTIONS ................................................................................................... 37 II.5.7 CNNS FOR MRI BRAIN TUMOR CLASSIFICATION ............................................................... 38 II.5.7.1 Input layer ......................................................................................... 39 II.5.7.2 Convolutional Layer ........................................................................... 39 II.5.7.3 Pooling Layer ..................................................................................... 40 II.5.7.4 Fully Connected Layer ........................................................................ 41 II.5.7.5 Logistic or Softmax Layer (LOSS) ......................................................... 41 II.5.7.6 Output Layer: The Final Decision Maker ............................................. 41 II.6 PERFORMANCE EVALUATION MEASUREMENTS ......................................................................... 42 II.6.1 CONFUSION MATRIX................................................................................................... 42 II.6.2 ACCURACY................................................................................................................ 42 II.6.3 SENSITIVITY .............................................................................................................. 42 II.6.4 SPECIFICITY ............................................................................................................... 43 II.6.5 F1-SCORE................................................................................................................. 43 II.6.6 ROC CURVE .............................................................................................................. 43 II.7 CONCLUSION ...................................................................................................................... 44 III. CHAPTER III BRAIN TUMOR CLASSIFICATION ...................................................................... 45 III.1 INTRODUCTION ................................................................................................................... 45 III.2 TAXONOMY OF MRI BRAIN IMAGE CLASSIFICATION ALGORITHMS ................................................ 45 III.3 CLASSIFICATION PROCESS ..................................................................................................... 48 III.3.1 PRE-PROCESSING STEP ................................................................................................ 49 III.3.2 FEATURE EXTRACTION STEP .......................................................................................... 49 III.3.2.1 Binarized Statistical Image Features (BSIF) ......................................... 50 III.3.2.2 Histogram of Oriented Gradients (HOG) ............................................. 51 III.3.2.3 GIST Descriptor: A Low-Dimensional Image Representation ............... 51 III.3.2.4 AlexNet .............................................................................................. 52 III.3.2.5 VGG-16 .............................................................................................. 52 III.3.2.6 Residual Network (ResNet-50 and ResNet-18) ................................ 53 III.3.3 CLASSIFICATION STEP: ................................................................................................. 53 III.3.3.1 Naive Bayes ....................................................................................... 54 III.3.3.2 K-Nearest Neighbor............................................................................ 54 III.3.3.3 Support Vector Machine (SVM) .......................................................... 55 III.3.3.4 Decision Trees .................................................................................... 57 III.4 OVERVIEW OF BRAIN TUMOR CLASSIFICATION .......................................................................... 58 III.4.1 MACHINE LEARNING TECHNIQUES ................................................................................. 58 III.4.2 DEEP LEARNING TECHNIQUES ....................................................................................... 59 III.5 DEEP RULE BASED CLASSIFIER FOR MRI BRAIN TUMOR CLASSIFICATION ........................................ 60 III.5.1 GENERAL ARCHITECTURE OF THE DRB CLASSIFIER ............................................................. 60 III.5.2 MASSIVELY PARALLEL FRB ........................................................................................... 62 III.5.2.1 Training process of the DRB system.................................................... 63 III.5.2.2 Validation process of the DRB system ................................................ 65 III.5.2.3 Decision Maker .................................................................................. 65 III.6 CONCLUSION ...................................................................................................................... 67 IV. CHAPTER IV: DRB-BBSIF FOR BRAIN TUMOR CLASSIFICATION ............................................. 68 IV.1 INTRODUCTION ................................................................................................................... 68 IV.2 THE ARCHITECTURE OF PROPOSED DRB-BBSIF CLASSIFIER ......................................................... 68 IV.2.1 EXTRACTION OF THE REGION OF INTEREST (ROI) .............................................................. 69 IV.2.2 EXPLORING BINARIZED STATISTICAL IMAGE FEATURES (BSIF) .............................................. 69 IV.2.3 DEEP RULE-BASED CLASSIFIER FOR MRI BRAIN TUMOR CLASSIFICATION ................................ 71 IV.3 EXPERIMENTS AND RESULTS .................................................................................................. 72 IV.3.1 DATABASE ................................................................................................................ 72 IV.3.2 EXPERIMENT 1 – CONSTRUCTION OF BANK OF BSIF FILTERS ............................................... 73 IV.3.2.1 Objectives and Methodology ............................................................. 73 IV.3.2.2 Analysis of Results.............................................................................. 74 IV.3.2.3 Interpretation of Results .................................................................... 76 IV.3.2.4 Key findings ....................................................................................... 78 IV.3.3 EXPERIMENT 2: IMPACT OF FEATURE EXTRACTOR METHODS ............................................... 78 IV.3.3.1 Objective of Experiment 2 .................................................................. 78 IV.3.3.2 Analysis of results .............................................................................. 78 IV.3.3.3 Interpretation of results ..................................................................... 79 IV.3.3.4 Key findings ....................................................................................... 80 IV.3.4 EXPERIMENT 3: EVALUATION OF THE DRB-BBSIF MODEL .................................................. 80 IV.3.4.1 Objective of Experiment 3 .................................................................. 80 IV.3.4.2 Analysis of Results.............................................................................. 80 IV.3.4.3 Interpretation of Results .................................................................... 82 IV.3.4.4 Key Finding ........................................................................................ 83 IV.4 CONCLUSION ...................................................................................................................... 83 V. CHAPTER V: DRB WITH DEEP FEATURE EXTRACTION ........................................................... 85 V.1 INTRODUCTION ................................................................................................................... 85 V.2 PROPOSED METHODOLOGY ................................................................................................... 85 V.2.1 PRE-PROCESSING STEP ................................................................................................ 86 V.2.2 FEATURE EXTRACTION STEP .......................................................................................... 86 V.2.3 CLASSIFICATION STEP .................................................................................................. 86 V.3 DATABASE ......................................................................................................................... 86 V.4 EXPERIMENTS AND RESULTS .................................................................................................. 87 V.4.1 EXPERIMENT 1: ALEXNET WITH 4 DIFFERENT CLASSIFIERS ................................................... 88 V.4.1.1 Analysis of Results.............................................................................. 89 V.4.1.2 Key Finding ........................................................................................ 91 V.4.2 EXPERIMENT 2: VGG-16 WITH 4 DIFFERENT CLASSIFIERS.................................................... 92 V.4.2.1 Analysis of results .............................................................................. 93 V.4.2.2 Key Findings ....................................................................................... 94 V.4.3 EXPERIMENT 3: RESNET-50 WITH 4 DIFFERENT CLASSIFIERS ................................................ 96 V.4.3.1 Analysis of Results.............................................................................. 97 V.4.3.2 Key Findings ....................................................................................... 98 V.4.4 EXPERIMENT 4: RESNET-18 WITH 4 DIFFERENT CLASSIFIERS .............................................. 100 V.4.4.1 Analysis of Results............................................................................ 100 V.4.4.2 Key findings ..................................................................................... 103 V.4.5 COMPREHENSIVE ANALYSIS OF RESULTS ........................................................................ 103 V.4.5.1 Classifiers: Strengths and Challenges ................................................ 104 V.4.5.2 Deep Learning Features ................................................................... 105 V.5 COMPARISON BETWEEN THE TWO CONTRIBUTIONS .................................................................. 108 V.6 HIGH PERFORMANCE OF THE DRB CLASSIFIER......................................................................... 109 V.6.1 CHARACTERISTICS OF DRB CLASSIFIER .......................................................................... 109 V.6.2 PERFORMANCE METRICS OF THE DRB CLASSIFIER ........................................................... 110 V.6.3 COMPARISON WITH OTHER CLASSIFIERS........................................................................ 110 V.7 CONCLUSION .................................................................................................................... 111 CONCLUSION AND PERSPECTIVES ............................................................................................. 112 Côte titre : DI/0090

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
DI/0090	DI/0090	Thèse	Bibliothéque des sciences	Anglais	Disponible Disponible

---

1 (1 - 1 / 1)