| Titre : | Brain Tumor Growth Modelling |
| Auteurs : | Kamel Adel, Auteur ; Rachida Saouli, Directeur de thèse |
| Type de document : | Thése doctorat |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2022 |
| Format : | 1 vol. (103p.) / couv. ill. en coul / 30 cm |
| Langues: | Français |
| Résumé : |
Prediction methods of Glioblastoma tumors growth constitute a hard task due to the lack of medical data, which is mostly related to the patients’ privacy, the cost of collecting a large medical dataset, and the availability of related notations by experts. In this thesis, we study and propose a Synthetic Medical Image Generator (SMIG) with the purpose of generating synthetic data based on Generative Adversarial Network in order to provide anonymized data. In addition, to predict the Glioblastoma multiform (GBM) tumor growth we developed a Tumor Growth Predictor (TGP) based on End to End Convolution Neural Network architecture that allows training on a public dataset from The Cancer Imaging Archive (TCIA), combined with the generated synthetic data. We also highlighted the impact of implicating a synthetic data generated using SMIG as a data augmentation tool. Despite small data size provided by TCIA dataset, the obtained results demonstrate valuable tumor growth prediction accuracy |
| Sommaire : |
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i ACKNOWLEDGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii LIST OF TERMS AND ABBREVIATIONS . . . . . . . . . . . . . . . . . . . xii List of Terms and Abbreviations 1 1 Medical background 6 1.1 Glioblastoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.2 GBM Investigations level . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.2.1 In Vivo: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.2.2 In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.2.3 In Silico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.3 Medical Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.3.1 MRI imaging techniques . . . . . . . . . . . . . . . . . . . . . . 16 1.3.2 MRI Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 1.3.3 Challenge in medical images: . . . . . . . . . . . . . . . . . . . . 19 1.4 Dataset (ADNI, BraTS, TCIA) . . . . . . . . . . . . . . . . . . . . . . . 21 1.4.1 The Cancer Imaging Archive (TCIA) . . . . . . . . . . . . . . . . 21 1.4.2 Alzheimer’s Disease Neuroimaging Initiative ADNI: . . . . . . . 22 1.4.3 Brain Tumor Segmentation BraTS . . . . . . . . . . . . . . . . . 23 1.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2 Tumor growth modeling 24 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.2 The math of GBM growth . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.2.1 Microscopic models .............................................................................25 2.2.2 Macroscopic models.............................................................................27 2.2.3 Overview of mathematical modeling approaches ................................31 2.3 Genetic of Brain Tumor ...................................................................................32 2.3.1 State of the art for Genetic investigation:.............................................33 2.4 Proposed MLP for genetic profile prediction:..................................................37 2.4.1 Materiel and methods:..........................................................................38 2.4.2 Data preparation...................................................................................39 2.4.3 Model Conception and Implementation :.............................................39 2.4.4 Our model classification results :.........................................................41 2.5 Discussion and Conclusion ..............................................................................42 3 Brain Tumor Growth Predictor 44 3.1 Motivation and objectives ................................................................................45 3.2 Background theory: machine learning .............................................................46 3.2.1 Machine learning for health care..........................................................48 3.3 Brain tumor growth prediction pipeline ...........................................................48 3.3.1 MRI image quality limitations .............................................................49 3.3.2 Skull stripping ......................................................................................50 3.3.3 Registration and resizing......................................................................52 3.3.4 Normalization and standardization ......................................................53 3.3.5 Filtering and denoising.........................................................................55 3.4 TGP model conception and architecture .....................................56 3.4.1 Training parameters .............................................................58 3.4.2 Implementation and working environment .........................59 3.5 TGP Experimental results ...........................................................59 3.5.1 Evaluating TGP performance...............................................................61 3.6 Discussion and Conclusion ......................................................63 4 Synthetic Medical Image Generator 65 4.1 Introduction ...........................................................65 4.2 Motivation and objectives ....................................................66 4.3 Background Theory : Generative adversarial networks (GANs).....................66 4.3.1 Generative adversarial networks: Model conception...........................66 4.3.2 Generative adversarial networks: Training strategy ............................67 4.3.3 Game theory ...................................................................68 4.4 State of the art: GANs in Medical investigation ..............................................69 4.4.1 Synthetic medical image generations:..................................................70 4.4.2 Medical image quality enhancing. .......................................................71 4.5 Synthetic Medical Image generator (SMIG) pipeline ......................................71 4.5.1 SMIG Model Training..........................................................................74 4.5.2 SMIG model Results ..........................................................75 4.5.3 model limitation and future work.........................................................76 4.5.4 Conclusion ............................................................77 5 Synthetic MRI images to predict GBM growth 78 5.1 Introduction .....................................................78 5.1.1 Problem statement................................................................................78 5.1.2 Objectives and motivation....................................................................78 5.2 Related work ...........................................................................79 5.3 SMIG data based on TCIA dataset...................................................................80 5.4 Tumor Growth Prediction Based on SMIG data..............................................81 5.5 Evaluating strategy..........................................................................83 5.5.1 Combination result validation..............................................................84 5.6 Experimental results................................................................86 5.7 Discussion .......................................................................................88 5.8 Conclusion..................................................................89 REFERENCES ............................................................................ 92 LIST OF PUBLICATIONS ..................................................................105 |
| En ligne : | http://thesis.univ-biskra.dz/5674/1/PhD_thesis.pdf |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| TINF/170 | Théses de doctorat | bibliothèque sciences exactes | Consultable |
