An Effective Approach for Sub-acute Ischemic Stroke Lesion Segmentation by Adopting Meta-Heuristics Feature Selection Technique Along with Hybrid Naive Bayes and Sample-Weighted Random Forest Classification

Abstract

In Earth, the most crucial illness for cause of death is ischemic stroke. Ischemic stroke arises as a result of an obstacle within a blood vessel supplying blood to the brain. In this paper, for sub-acute ischemic stroke lesion segmentation, we utilize an effective meta-heuristic feature selection technique along with hybrid Naive Bayes (NB) and sample weighted random forest (SWRF) classification approach. Initially, the features are extracted from the pre-processed image, after that, the feature selection is done by using the multi-objective enhanced firefly algorithm. To improve the classification performance, the dimensionality of the feature vectors and errors are reduced by eliminating such irrelevant and redundant features. After the feature selection process, an ensemble of NB and SWRF classifiers is used for segmenting the image. Here the NB classifier is trained and applied to estimate the weights of training samples. Then, the training samples with estimated weights are utilized to train SWRF. In our work stroke lesion segmentation is formulated as a binary classification problem where every local region is classified as either affected or non-affected area.