The sagittal Sinus running down the center of the brain provided a challenge for us when we injected into the 3rd ventricle in our hamster model.  We actually bored our cranial hole lateral to the vessel to avoid bleeding and and lowered the cannula.  Once the tip was below the vessel, we then moved the cannula back to zero (or ML coordinate) to make midline injection. The idea was to push the vessel over and not damage it.  The rest was stereotaxic surgery according to our map coordinates.  I'm not sure if this is helpful for your surgery, but my experience. While thinking about your particular issue with the virus, I might also add, that if you are incorrect in your placement or leaving a needle track to the third ventricle, you could be loosing your virus to to the CSF pool.  You might want to make a small pocket below your injection site (still within your nuclei of interest) and slow down your infusion rate.  Just put your needle down to the new desired location and then back it off (bring it up a few mm in the z direction) and then using your pump inject over a few minutes.  I have had luck with that technique when injecting glioma cells as they would often go the path of least resistance if they weren't allowed to diffuse in and penetrate the discreet area we wanted them to grow in.  Also worked for microspheres, drugs, and lesion cytotoxins.

Best of luck