Computer vision based identification of abnormal tissues in biomedical images

Easy visualization and recognition of slight abnormality in the body of the human can be done through various methods of biomedical imaging. Abnormalities might be due to presence of tumor, also known as the group of abnormal cells that can directly destroy all healthy cells. In case of Brain these abnormal cells grows inside or around the brain. These abnormalities turn destructive and plays a determinative role in the quality of the health of the human and thus increasing the life expectancy and longevity. In early times the diagnosis of the tumors in brain was exhausting task as the symptoms that can be detected physically can only be seen in the advance stages of the tumor. In modern times imaging methods like Magnetic Resonance Imaging (MRI) provides efficient and meticulous insight of tumor condition. It supports the treatment at preliminary stage. In Digital Imaging and Communications in Medicines (DICOM) images, implementation of the image processing techniques help in the detection of the most minute cell with less probability of human error, better speed and high efficiency. Here the identification of abnormal tissue in biomedical images based on computer vision has been used. The features on which the abnormal and the normal images are differentiated are namely area, perimeter and entropy. Entropy has been extracted using the feature extraction methodology from Gray Level Co-occurrence Matrix (GLCM) of the sampled of tumor image.