Author Email: email@example.com
Brain tumor is a serious disease and numbers of affected people are drastically increasing day by day and researchers are focused on detection and classification of such types of severe diseases so that appropriate diagnosis can be provided to patient. In medical science, manually detection of brain tumor using medical resonance image by radiologist or neurosurgeon is tedious and time consuming procedure. One can save the time and energy of radiologist/ neurosurgeon by proving a framework which will automatically and reliably classify and detect the grade of tumor. The aim of this paper is to present a structure which not only classify the tumor as benign or malignant but also specify the grade of tumor. This framework comprises of different image processing and machine learning algorithm like pre-processing, segmentation, feature extraction and classification. At the end of this paper algorithm wise outcome is presented and evaluation of framework is carried out using different evaluation measures.
In medical science, Medical Resonance Image (MRI) is one of the widely used imaging techniques which is reliable and safe in order to detect the brain tumor. MRI is carried out using a magnetic field and radio waves to create a complete image of the brain which can further used to test different conditions of brain like swelling, tumor, cysts, infection and bleeding conditions. MRI is extensively used by analysts to assess the tumors   but huge amount of data produced by such imaging technique prevents manual segmentation. Manual segmentation of such images is time consuming, tedious and prevents the use of precise quantitative measurements. In diagnosis of tumor, the precise, accurate detection of size & volume of brain tumor is more important in addition to its location. It’s the need of the hour to make use of computer-aided tools in medical science so that aforementioned phenomenon can be carried out automatically and reliably.
A framework for the brain tumor detection is described in this paper is efficient and fully automated and useful in medical applications especially for radiologist and neurosurgeon. Prominent techniques of image processing and machine learning are employed in creation of this framework. Experiments on different images prove the validity of system. An important scope would be use of T2 axial MRI images and other evaluation metric for evaluation purpose.
210 total views, 1 views today