Shubha76/Brain-Tumor-Detection-from-MRI-images-Spring-2018

Earlier detection of brain tumors plays a vital role in its treatment as well as dynamically increase the survival rate of the patients. Magnetic Resonance Imaging (MRI) scans are widely used to diagnose the brain tumors which provides better accuracy than other medical imaging techniques. Still, the manual segmentation of MRI images and detecti...

Earlier detection of brain tumors plays a vital role in its treatment as well as dynamically increase the survival rate of the patients. Magnetic Resonance Imaging (MRI) scans are widely used to diagnose the brain tumors which provides better accuracy than other medical imaging techniques. Still, the manual segmentation of MRI images and detecting the brain tumors is a time consuming and prone to error task, which is currently done by the medical experts or radiologists. So, there is an evident necessity for automatic brain tumor segmentation and extracting various characteristics of brain tumors. In this study, three widely used standard image segmentation methods (threshold based, k-means clustering and watershed segmentation) has been tested using collected brain MRI images to isolate the tumors from the rest of the brain regions, and their performance was compared based on the segmentation output. K-means clustering showed a better result than two other methods. Besides this, a graphical user interface (GUI) is designed based on primary image processing techniques and by using the solidity feature of brain tumors. Two of the highly useful brain tumor characteristics (area, and perimeter) are also measured here and displayed on the output window of GUI. The accuracy of this application for tumor detection on brain MRI images and features calculation is much high. More features can be extracted, and the accuracy can be maximized by following some other rigorous techniques, which later could be highly helpful for the medical practitioners working in this field.