Introduction/purpose The benefits of learning on simulator have long been established across different areas of practice including medical industry. In particular recent technological advances have produced either flow models or simulators on which practitioners can learn new techniques evaluate new devices or improve their skills outside of direct patient care. None of the models available do however reproduce entirely the complex anatomy and variants encountered in real medical practice. Moreover the use of such models in an MRI environment is prohibitive.

The purpose of this experiment was to create a practical cadaveric flow model for specific use in the MRI environment.

Materials/method The current presented work was carried at the IMSAT MRI laboratory, University of Dundee, UK. A total of three Thiel cadavers were used for trans ventricular access in the ascending aorta and further in the cerebro-vascular circulation. The access followed a 5 cm incision over the left sternal border. Using classical Seldinger technique the left ventricle cavity was thereafter accessed and using fluoroscopy an Amplatz wire was passed into the ascending aorta. A 5F catheter was exchanged and navigated into the ascending aorta. Intravascular position of the catheter was documented by injecting a small amount of contrast in the ascending aorta. The Amplatz wire was navigated into the proximal descending aorta and by a series of dilators a, 16F endotracheal tube was navigated into the ascending aorta and secured by inflating the endotracheal balloon.

Using fluoroscopy, catheters were navigated into the left, right common carotid and right vertebral arteries. The catheters were connected to a surgical heart pump with control for heart rhythm and output.

In one case a direct right jugular access into the superior sagittal sinus was performed and a 7F balloon catheter was connected to a suctioning component of the heart pump in order to reproduce the right heart function and avoid brain swelling.

Results In all three cases the trans ventricular access was established and navigation into the cerebro-vascular vessels was possible and confirmed by fluoroscopy. The fluoroscopic and MRI visualization of cerebral vasculature was possible to the level of MCA bifurcation and proximal M2 segments, for the anterior circulation and through the entire posterior circulation to the level of distal P1. Perfusion with a continuous small amount of fluid was possible for a limited period of time; brain swelling was noticed after prolonged use in all cadavers.

The use of a separate suctioning pump for the superior sagittal sinus did not prevent brain swelling and CSF extravazation of the infused fluid.

Conclusion Reproducible development of a cadaveric cerebrovascular model is possible. The main purpose of such model is the use in an MRI environment for establishing intra-operative optimal sequences and instrument development and navigation. Current limitations of the model are related to the lack of integrity of the blood-brain barrier and subsequent interstitial and subarachnoid extravazation.

Disclosures B. Cox: None. R. Buciuc: None. M. Rube: None. A. Melzer: None.