Fetal head localization and fetal brain segmentation from MRI using the center of gravity

Magnetic resonance imaging (MRI) of fetus is an essential tool to study normal and abnormal brain development of fetus. A meaningful interpretation of fetal brain by a neurologist requires brain segmentation. We propose an automatic method for segmentation of the fetal brain in MRI. We first detect the fetal head location in 2D MRI slices using center of gravity (COG), and draw a circle on the input image that encloses the fetal head and is taken as the ROI. The fetal brain is segmented from the ROI using thresholding, morphological operations and connected component analysis. Our method is evaluated on 25 volumes of fetal data. To assess the performance of the method we compute the Dice similarity coefficient (D) and obtained the value as 0.8842 which is comparable to all the existing methods.