dc.contributor.author |
Abdalla, Hussna Elnoor Mohammed |
|
dc.contributor.author |
Supervisor - Mohamed Yagoub Esmail |
|
dc.date.accessioned |
2015-11-10T09:58:25Z |
|
dc.date.available |
2015-11-10T09:58:25Z |
|
dc.date.issued |
2015-08-27 |
|
dc.identifier.citation |
Abdalla,Hussna Elnoor Mohammed . Brain Tumor Detection by Using Artificial Neural Networks \ Hussna Elnoor Mohammed Abdalla ;Mohamed Yagoub Esmail .-Khartoum : Sudan University of Science and Technology,Engineering ,2015 .-101 .:ill.;28 cm .- M.sc |
en_US |
dc.identifier.uri |
http://repository.sustech.edu/handle/123456789/11813 |
|
dc.description |
Thesis |
en_US |
dc.description.abstract |
Brain tumor is one of the most dangerous diseases which require early and accurately detection methods, current used detection and diagnosis methods for image evaluation depend on decision of neuro-specialists, and radiologist which possible to human errors. Manual classification of brain tumor is time consuming. This study describes the processes and techniques used in detecting brain tumor from magnetic resonance imaging (MRI) and ANN techniques, which are of the most application of artificial intelligent that used in biomedical image classification and recognition.
In the proposed system, features are extracted from raw MRI images which are then fed to ANN through GUI to received suspected MRI for early tumor detection. This thesisimplemented in the different stages for Computer Aided Detection System (CAD-system) after collected the image data (magnetic resonance images); first stage is pre-processing and post-processing of MRI images to enhancement it and then the processed image is being more suitable to analysis. The study was used threshold to segment the MRI images by applied mean gray level method. A comprehensive feature set , computed and ANN rules are selected to classify normal and abnormal image.
In the second stagestatistical feature analysis was used to extract features from MRI images; the features were computed using Haralick’sequation for feature based on the spatial gray level dependency matrix (SGLD).
The suitable and best features to detect the tumor in image were selected. In the third stage the artificial neural networks (ANN)were designed; the feed-forward back propagation neural network with supervised learning were apply as automatic method to classify the images under investigation into tumor or none tumor . The network performances were evaluated successfully tested and achieved best results with accuracy of 99%,specificity100% and sensitivity 97.9%. |
en_US |
dc.description.sponsorship |
Sudan University of Science and Technology |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
Sudan University of Science and Technology |
en_US |
dc.subject |
Biomedical Engineering |
en_US |
dc.subject |
Brain Tumor |
en_US |
dc.subject |
Neural Networks |
en_US |
dc.subject |
Using Artificial |
en_US |
dc.title |
Brain Tumor Detection by Using Artificial Neural Networks |
en_US |
dc.title.alternative |
الكشف عن أورام الدماغ باستخدام الشبكات العصبية الاصطناعية |
en_US |
dc.type |
Thesis |
en_US |