Tumor Segmentation from Brain MRI using Clustering Algorithms

Moksh Shukla
Mar 6, 2020 Project

Duration: July 2019 - March 2020

Mentor: Prof. Krishna K. Sharma, Dept. of Computer Science & Informatics, University of Kota

Details

  • Compared various Noise Filters on images with Gaussian Noise using PSNR as the test parameter
  • Implemented Contour-Detection Based Cropping of MRI images as a pre-processing step which improved accuracy of clustering algorithms and reduced their computation time by clipping unwanted portion from MRI
  • Tested 9 different combinations of Noise Filters & Clustering Algorithms for their tumor classification accuracy, with Anisotropic Diffusion Filter along with Fuzzy K-Means achieving the best Classification Accuracy of 94.55%
  • Devised an efficient algorithm for Tumor Detection involving: Contour-based Cropping, De-Noising using Anisotropic Diffusion Filter, Segmentation with Fuzzy K-Means & Post-Processing with Erosion & Median Filter
Machine Learning Clustering Research

Publications

A Comparitive Study to Detect Tumor in Brain MRI Images using Clustering Algorithms

The identification of the tumor area in the magnetic resonance images(MRI) by radiologists or experts is a tedious and time-consuming …
PDF Project Source Document DOI