Loading...

Diagnosis and Prediction of Coronary Arteries Disease by Applying Data Mining and Image Processing Techniques

Hasoni Shahre Babak, Mohammad Sagegh | 2019

698 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 51943 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Khedmati, Majid; Foroozan Nia, Khalil
  7. Abstract:
  8. Heart disease is one of the major causes of death in all countries, especially developing countries. At the moment, using Image Processing methods as well as analysis of electrocardiographic signals, heart disease is diagnosed with the help of specialists. Applying artificial intelligence and machine learning methods, many studies attempted to provide models that are used to diagnose automatically the heart disease without the need for a specialist and only relying on the past data. But less is done on CTA images of the heart. Hence, in this thesis, a new method for image processing and a Multi Support Vector Machine (MSVM) classification for coronary artery disease detection based on CTA images is presented. In this method, feature extraction with two methods of pulmonary implantation and unsupervised learning is done in the Autoencoder artificial neural network and then helps MSVM to be trained to diagnose the disease. By testing the proposed method on several datasets, including a dataset from the Tehran heart center, as well as a heart dataset from the UCI website, the results show that, firstly, the MSVM method compared to SVM and other existing algorithms such as KNN and NB and MLP have a higher accuracy in a set of standard data such as Heart, and secondly, the proposed method for extracting features with Autoencoder is very efficient. So that even with a small number of training samples it could produce an Accuracy more than 88% and Precision and Recall about 90%
  9. Keywords:
  10. Coronary Arteries Disease (CAD) ; CT Angiography (CTA) ; Machine Learning ; Multisupport Vector Machin ; Image Processing ; Support Vector Machine (SVM)

 Digital Object List

 Bookmark

No TOC