Segmentation of needle artifacts in real time 2D MR images

During percutaneous minimally invasive procedures a needle is used to access the region to be treated without need for an open surgery. Remarkable is the increasing use of percutaneous thermal ablation to treat neoplastic formation. The effectiveness of such procedures is highly dependent on the correct placement of the needle inside the region to be treated. Imaging monitoring provides the physician with the possibility to inspect the location of the device, which is responsible for a signal void in the MR images acquired referred to as needle artifact. The procedure can be performed under real time Magnetic Resonance Imaging guidance. In this work two algorithms for automatic needle detection in real time MR images were developed. The detection is anticipated to increase the accuracy of the device positioning