Alpha Beta Pruned UNet - A Modified UNet Framework to Segment MRI Brain Image to Analyse the Effects of CNTNAP2 Gene towards Autism Detection

Abstract

The brainchild of the proposed work lies in automatic detection of autism using image segmentation method. CNN is the most powerful technique for biomedical image segmentation where several variants are proposed. The proposed work mainly focuses on modified UNet segmentation method for image segmentation and classification called Alpha-Beta Pruned UNet which is a dimensionality reduction technique in UNet. A comparison metric is also made between UNet and the proposed algorithm with the experimental results.