ABSTRACT
Today’s era is of big and complex data. In the fields of medicine, science, engineering, commerce, sports, etc., large amounts of data sets are produced and processed daily. Such data require analysis to extract meaningful insights without which such data is a complete waste. Data provide vigor to the researcher to research and find something interesting and helpful for society. Healthcare data/images are generated in large amounts, and it is becoming more complex. Based on proper data available, primarily for brain tumor data, this proposed work will be able to make proper healthcare analysis for preventive and remedial decisions.
Abnormal, uncontrolled growth of cells in brain is known as “brain tumor”. Its detection and diagnosis is an active area of research in brain image processing, as diagnosis of brain tumor is a serious concern.
Medical imaging is the best method for diagnosis of many diseases. It also plays important role in diagnosis of brain tumor. Medical experts take time to study and analyze brain images to conclude the presence of brain tumor, as various tumors having different size, shape and structure. Therefore, automated expert system, which can diagnose the brain tumor will be more beneficial. Such automated systems will save diagnosis time and gives accurate diagnosis, without the need of interference of medical expert. Magnetic resonance images (MRI) and computed tomography (CT) can be used to scan the tumor and diagnose it. In MRI, it is challenging to find location and size of the tumor. Images contain high directional features; many can be relevant or irrelevant. Irrelevant features can degrade performance of automated expert system.
To improve accuracy of the automated system, relevant features should be selected from images. Feature selection is one of the complex tasks. Many complex problems can be solved very easily using nature inspired algorithms. Nature inspired algorithm has widely been used in many literatures for feature selection.
This chapter discusses about diagnosing of brain tumor using nature inspired computing. Evolutionary and swarm intelligence are population-based optimization cum prevalent optimization algorithms inspired by nature. The aim of this chapter is to present how to apply swarm intelligence and evolutionary algorithm for diagnosis of brain tumors.
