An Evaluation of Automated Tumor Detection Techniques of Brain Magnetic Resonance Imaging (MRI)

Abstract

Image processing is a technique developed by computer and Information technology scientist and being used in all field of research including medical sciences. The focus of this paper is the use of image processing in tumor detection from the brain Magnetic Resonance Imaging (MRI). For the brain tumor detection, Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are the prominent imaging techniques, but most of the experts prefer MRI over CT. The traditional method of tumor detection in MRI images is a manual inspection which provides variations in the results when analyzed by different experts, therefore, in view of the limitations of the manual analysis of MRI, there is a need for an automated system that can produce globally acceptable and accurate results. There is enough amount of published literature available to replace the manual inspection process of MRI images with the digital computer system using image processing techniques. In this paper, we have provided a review of digital image processing techniques in the context of brain MRI processing and critically analyzed them for the identification of the gaps and limitations of the techniques so that the gaps can be filled and limitations of various techniques can be improved for precise and better results.

Using Probabilistic Classification Technique and Statistical Features for Brain Magnetic Resonance Imaging (MRI) Classification: An Application of AI Technique in Bio-Science

“…mostly the skull is removed from the brain MRI which contains some components which do not represent the actual brain [25]. Actually, the brain constitutes a small portion of the whole brain MRI.…”

An Offline Signature Verification Technique Using Pixels Intensity Levels

“…The Intrinsic information can be given by (8). The Gini Index is an alternative of measuring information gain and can be given by (9), it measures impurity instead of entropy. The average Gini Index can be given by (10), it is used instead of an average entropy of information gain.…”

Critical Evaluation of Frontal Image-Based Gender Classification Techniques

“…Humans can also identify the gender accurately although if the male or female face is scratched. The gender classification is a key problem for the machines like MRI Classification, Signature Verification and Pattern recognition problems [2][3]. The computerized gender classification is gaining much popularity because of rapid development in the fields of the internet, human-computer interaction, Introduction of an electronic system in banking and commerce, security [4][5].…”