Locating Brain Abnormality and Extracting Features from MRI CT Images using Segmentation and Classification Techniques

Abstract

The ever challenging brain abnormalities diagnosis by medical professionals based on the data from MRI and CT scans needs deeper and finer information to strengthen diagnosis and subsequent treatment. So for on this track several researchers have contributed to give possible information for diagnostic however there are certain areas where the accuracy needed to be improved for diagnosis and subsequent effective treatments. The identified challenges needing focused attention enhancing extraction, segmentation accuracy consideration of majority of features of the images and giving a support of single platform for multi abnormalities diagnosis. newlineThe proposed work undertaken in this research improves segmentation accuracy taking in combination symmetry and textures features followed by categorized recognition of brain images. The analysis of categorized data implementing data mining and machine learning algorithms leads classification at better level and detection accuracy at higher level. The suggested track of processes sequences and normally supported under different frameworks and hence to be collected in a single window from different frameworks methodology output. This issues are resolved in the proposed framework integrating scattered processes in a sequence and utilizing the output of one into the next in sequence has justified the targeted accuracies and enhancement levels. The implementation of proposed work could achieved improved segmentation accuracy evaluated using segmentation performance evaluators viz PRI as 0.9857, VOI as 0.022, GCE as 0.01424. Also, the classifier accuracy were evaluated by performance measures viz Accuracy, Specificity, Sensitivity, F-measure, Precision and Recall which depicted as the research effectively demonstrates various brain hemorrhages as Epidural hemorrhage (EDH), subdural hemorrhage (SDH), subarachnoid hemorrhage (SAH), Intracerebral hemorrhage (ICH) and Intraventricular hemorrhage (IVH) by its dimension, shape and area in the brain which can use for accurately detecting