[Advances in the contributions of imaging to stereotaxic localization of cerebral arteriovenous malformations for radiosurgery ].

Historically, angiography was one of the first diagnostic methods to allow for visualization of neurovascular structures. It has been and still is very useful for precise evaluation of vascular pathology and is one of the main elements in treatment planning for radiosurgical targets. It is the only imaging method that gives insight into the angioarchitecture of a cerebral arteriovenous malformation, possibly reducing the target volume. Construction of frames (Leksell, Fisher) that are compatible with cross-sectional imaging methods, such as CT and MRI allowed there use for planning of stereotactical treatment for brain cerebral arteriovenous malformations. The advantages of these methods are given by the fact, that they are less invasive and that they allow visualization of neurovascular structures and surrounding cerebral structures. Further evolution of the cross-sectional imaging techniques allowed reconstruction of the image data in different planes and segmentation of structures such as vessels. Use of special algorithms allow visualization of the image data, i.e. surface rendering with 3D images of vascular structures. However, such images allow no detailed insight into the angioarchitecture of a cerebral arteriovenous malformation and give rather a view of the whole volume, i.e. a "tumor" aspect of the cerebral arteriovenous malformation. Similar images are currently also obtained with digital substraction angiography using rotational image acquisitions and image postprocessing allowing 3D reconstruction of angiographical image data. The different image evaluation methods are thus complementary all giving useful information for treatment planning. Therefore it would be useful to develop the possibility to integrate the information obtained by these modalities. Image fusion require identification of fiducial marks, what can be performed with application of external marks or by using internal anatomical marks. Recent developments allow now use of vascular structures as fidiucial marks to obtain image fusion. This paper reports on the evolution of stereotactical planning, performed on 541 patients over a period of eight years.