Non-rigid Cerebral Surface Registration for Neonates Using ICP Algorithm

Abstract

Magnetic resonance (MR) images are widely used to diagnose cerebral diseases. MR image registration based brain shape evaluation is effective to diagnose brain diseases because the diseases may deform the brain shape, and the deformed region differs among types of diseases. This paper focuses on neonatal brain MR images, and introduces a non-rigid image registration method using sulcal-distribution index (SDI), which is calculated from MR signal on the cerebral surface. The method is based on an iterative closest point (ICP) algorithm. The proposed method will be effective for the neonatal brain because the method evaluates the correspondence of cerebral sulci distribution. Results in seven neonates (modified age was between 3 weeks and 2 years) showed that the method registered the cerebral shapes successfully.