| Experimental Investigation of Recirculation Zone Characteristics of a Pre-filming Airblast Injector |
| Paper ID : 1008-ASAT19-FULL (R2) |
| Authors: |
|
Mohamed Nady Darwish * Master Student |
| Abstract: |
| Breast cancer is considered one of the most popular forms of cancer that is the top cancer in women. Regular checkups and tests are very important to early diagnosis for higher survival rate, but on the other side, the manual evaluation of the existence of such malignant tissue regions in images is a critical task, which can be assisted by new intelligent computerized technologies. Deep Learning is a leading technology for predicting tumors as malignant or benign tumors from histopathology images. Deep learning based techniques involve two main models: Convolutional Neural Network (CNN) model and Long Short Term Memory (LSTM) model. The former can extract important features from the internal representation of the image, while the latter can selectively remember patterns for a long duration of time. This paper presents different deep learning based approaches: CNN-based and hybrid combinations of LSTM- and CNN-based approaches, for breast cancer detection. These approaches are preceded with a segmentation layer for checking images of different sizes, and the whole structural framework for each approach is described and illustrated, separately. In paper, the proposed approaches are learned based on different standard data sets, evaluated among each other in several extensive experiments, and a comparative analysis with recent state-of-the-art approaches is listed. |
| Keywords: |
| Central Toroidal Recirculation Zone (CTRZ) |
| Status : Conditional Accept (Oral Presentation) |
