A fiducial-based framework for precise MRI-guided stereotaxic targeting in nonhuman primates
Year: 2026
Type: preprint
Abstract: ABSTRACT Background Accurate MRI-guided stereotaxic targeting in nonhuman primates requires reliable registration between MRI coordinates and stereotaxic space. We evaluated skull-fixed MR-compatible screw fiducials as a practical reference for this registration. Results Using a 3D-printed skull model with 28 fiducials and three internal targets, we compared offset translation, rigid registration, and affine registration under aligned and misaligned conditions. Offset translation was most sensitive to misalignment. Affine registration fit skull-surface fiducials closely, but rigid registration produced lower overall target errors. In vivo validation showed that implanted fiducials could be identified on MRI and re-measured during a later surgery. Comparison with existing methods This approach builds on previous fiducial-based stereotaxic methods by systematically comparing registration transforms, testing misalignment, evaluating fiducial number, and measuring accuracy at independent internal targets. Conclusions Skull-fixed fiducials provide a practical framework for MRI-to-stereotaxic registration. Rigid registration may offer the safest balance between correcting positioning differences and preserving accuracy at internal targets. Highlights MR-compatible screws provide an MRI-to-stereotaxic registration framework Offset translation is sensitive to stereotaxic misalignment Rigid registration improves accuracy at independent internal targets Affine registration can overfit skull-surface fiducials Implanted screw fiducials supported in vivo stereotaxic registration
Authors: Peyton Harmon, Reza Azadi
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7208713592