<p>Medical imaging and analysis plays a crucial role in diagnosis and treatment planning. The anatomical complexity of human brain makes the process of imaging and analyzing very difficult. In spite of huge advancements in medical imaging procedures, accurate segmentation and classification of brain abnormalities remains a challenging and daunting task. This challenge is more visible in the case of brain tumors because of different possible shapes of tumors, locations and image intensities of different types of tumors. In this paper we have presented a method for automated segmentation of brain tumors from magnetic resonance images. An enhanced and modified Gaussian mixture mode model and the independent component analysis segmentation approach has been employed for segmenting brain tumors in magnetic resonance images. The results of segmentation are validated with the help of segmentation evaluation parameters.</p>
Automated measurement of arterial input function in first-pass myocardial perfusion magnetic resonance images using independent component analysis
