NEURAL-NETWORK SEGMENTATION OF MAGNETIC-RESONANCE SPIN-ECHO IMAGES OFTHE BRAIN

This paper describes a neural network system to segment magnetic reson ance (MR) spin echo images of the brain. Our approach relies on the an alysis of MR signal decay and on anatomical knowledge; the system proc esses two early echoes of a standard multislice sequence. Three main s ubsystems can be distinguished. The first implements a model of MR sig nal decay; it synthesizes a four-echo multiecho sequence, in order to add images characterized by long echo-times to the input sequence. The second subsystem exploits a priori anatomical knowledge by producing an image, in which pixels belonging to brain parenchyma are highlighte d. Such anatomical information allows the following submodule to disti nguish biologically different tissues with similar water content, and hence similar appearance, which might produce misclassifications. The grey levels of the reconstructed sequence and the output of the second module are processed by the third subsystem, which performs the segme ntation of the sequence. Each pixel is assigned to one of five differe nt tissue classes that can be revealed with brain MR spin echo imaging . With a suitable encoding, a five-level segmented image can then be p roduced. The system is based on feed-forward networks trained with the back-propagation algorithm; experiments to assess its performance hav e been carried out on both simulated and clinical images.