W. Niessen and M. Viergever (Eds.): MICCAI 2001, LNCS 2208, pp. 1421-1423, 2001. Springer-Verlag Berlin Heidelberg 2001 The Clinical Use of Multi-modal Resources (2D/3D/Statistics) for Robot Assisted Functional Neurosurgery Alim-Louis Benabid 1 , Dominque Hoffmann 1 , Luc Court 1 , Vincent Robert 2 , SØbastien Burtin 2 ,Patrick Pittet 2 , Jrg Fischer 3 1 Service de NeuroChirurgie, Centre Hospitalier Universitaire, Grenoble, France 2 Integrated Surgical Systems SA, Bron, France, 3 IVS Solutions AG, Chemnitz, Germany Abstract. This paper presents specific techniques and their implementation to integrate in the planing of functional neurosurgery all the available resources (3D preoperative examinations, intra-operative 2D radiographic X-rays, statisti- cal database and atlases). Keywords. Functional Neurosurgery Image fusion Co-Registration Sta- tistical Data Stereotactic Robot. 1 Introduction We present in this paper different methods to combine all the available information for planning of stereotactic functional neurosurgery and their implementation within a planning software (VoXim™, IVS Solutions AG) . This software is used to drive a stereotactic robot (NeuroMate™, ISS SA) for an optimal targeting of the planned trajectory[2][3]. 1 2 Co-registration and Fusion Co-registration of two modalities consists in the computation of the rigid transforma- tion matrix im2 T im1 used to convert coordinates expressed in the first image referential in coordinates expressed in the second image referential. After co-registration of two modalities, it is possible to display a point defined on one modality on the other mo- dality and vice-versa. Stereotactic trajectories can also been displayed on both mo- dalities. Fusion consists in the construction of a new hybrid data set from co-registred exams. Fusion allows a very synthetic management of multi-modalities. For stereotactic applications, Fusion does not bring any additional information to co-registered modalities. 1 This work has been granted by the French Government