Brain portion segmentation from Magnetic Resonance Images(MRI) of human head scan using Richardson Lucy deconvolution and intensity thresholding

This article suggests a new scheme to extract brain portion from T1-W Coronal Magnetic Resonance Images (MRI) of human head scans. We propose that the Richardson-Lucy (RL) deconvolution algorithm can be employed to improve the boundary detection. Gaussian type point spread function(PSF) is assumed for the RL algorithm. The improved image is then subjected to binarization, morphological erosion and dilation, largest connected area to isolate the brain portion. Experiments with this scheme on 12 volumes of dataset collected from Internet Brain Segmentation Repository(IBSR) show that it performs better than the widely used Brain Surface Extractor(BSE) and Brain Extraction Tool(BET) methods.