Performance comparison of SVM, NB for MRI pancreas image classification

• Autores: B Aruna Devi, M Pallikonda rajasekaran
• Localización: Sustainable development in engineering and technology, 2022, ISBN 978-84-124943-4-1, págs. 337-352
• Idioma: inglés
• Enlaces
  • Texto Completo Libro (pdf)
• Resumen

  • Computer Aided Diagnosis (CAD) is an automatic route of diagnosing the tumourin medical field. Magnetic Resonance Imaging (MRI) has been normally applied inthe identification of tumour in our body. MRI is used for soft tissue contrast and noninvasiveness in medical image system. The troubles of MRI are incorrect diagnosisand large amount of time consuming by a radiologist. Programmed classifiers canmostly renew the detection process, in normal and abnormal images, naturally. Thisresearch work represents a smart, correct, and powerful approach to discriminatehuman pancreas magnetic resonance images (MRI) as normal or abnormal image.It presents the response of Naive bayes (NB) and Support vector machine (SVM)approaches on pancreas tumor classification. For that, we extract features from pancreas165 MR images applying GLCM approach and analysed by two classificationapproaches such as NB and SVM. NB approach has high classification accuracy (98%)which is higher than SVM.