الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 131 |  |  |
| | | from | 131 | | |
Abstract
Breast cancer is a gigantic burden on humanity causing loss of enormous numbers of lives and money. It is the world’s leading type of cancer among women and a leading cause of mortality and morbidity. The histopathological examination of the breast tissue biopsies is the gold standard for diagnosis. In this thesis, a Computer-Aided Diagnosis (CAD) system based on deep learning is developed to ease the pathologist’s mission. For this target, 5 pre-trained Convolutional Neural Network (CNN) models are analyzed and tested, which are: Xception, DenseNet201, InceptionResNetV2, VGG19, and ResNet152 with the help of data augmentation techniques and a new transfer learning approach. These models are trained and tested with histopathological images obtained from the BreakHis dataset. Multiple experiments are preformed to analyze the performance of these models through carrying out magnification dependent and magnification independent binary and eight-class classifications. Xception model has shown a promising performance through achieving the highest classification accuracies for all the experiments. It has achieved a range of classification accuracies from 93.32% to 98.99% for magnification independent experiments and from 90.22% to 100% for magnification dependent experiments.
This thesis offers many contributions in the field of breast cancer classification using histopathological images. The adopted models in this thesis execute testing on samples classified according to the variable magnification or independent of it. Additionally, it expresses the results as either binary classification or eight-class classification. Moreover, a single most efficient model is aimed by this thesis to be used in all the experiments and be generated as the prime diagnostic model for the pathologist. Finally, many models are used per experiment to evaluate the potential of accomplishing various successful models in a single experiment.