2026 Poster Presentations
P345: FROM CONCEPT TO CODE: LEVERAGING VIBE CODING AND AI TO BUILD MORPHONEURO, A NOVEL OPEN-SOURCE PLATFORM FOR NEUROSURGICAL RESEARCH
Filippo Sinosi, MD; Chiara Angelini, MD; Marco Obersnel, MD; Hao Tang, MD; Roberto Rodriguez Rubio, MD; UCSF
Objective: The creation of computational tools for neuroanatomical visualization and surgical planning is often hindered by surgeons’ limited programming expertise. Traditional software development requires collaboration with professional engineers, which slows translation from research to practice. With the emergence of vibe coding—an AI-driven workflow in which natural language prompts generate and refine functional code—clinicians can now directly participate in building software. This study presents MorphoNeuro, a 3D morphometric analysis tool for neurosurgical corridors, developed entirely through vibe coding with the assistance of AI models such as Claude (Anthropic, San Francisco) and supported by collaborative platforms such as GitHub (San Francisco).
Methods: Using Python and Blender as the base environment, neurosurgical researchers employed AI-assisted coding tools to iteratively design and debug code. Natural language prompts guided feature generation, while version control and community access were managed via GitHub. Key functionalities targeted included volumetric surgical corridor reconstruction, automated calculation of morphometric parameters (angles of attack, surgical windows, exposure volume), and standardized visualization for comparative studies. The workflow prioritized accessibility, allowing individuals without formal computer science training to act as primary developers.
Results: MorphoNeuro was implemented successfully on Windows, macOS, and Linux operating systems. Testing confirmed seamless import of stereotactic coordinates, construction of volumetric corridors, and automatic metric extraction. Additional features—such as automated ray-casting corridors, batch analysis—were integrated through further vibe coding iterations. All development was documented and shared openly on GitHub, ensuring reproducibility and fostering global collaboration. The process demonstrated that AI-assisted coding environments such as Claude can lower the technical threshold for software creation, allowing sugerons to generate tailored research tools from scratch.
Conclusion: The development of MorphoNeuro highlights how vibe coding, combined with AI assistants and open-source platforms, democratizes software innovation in neurosurgery. By empowering researchers to act as tool creators rather than passive users, this approach accelerates translational research and reduces dependence on commercial or proprietary systems. As the ecosystem of vibe coding tools continues to expand, more clinicians will gain the ability to design, refine, and share novel applications aligned with their research and surgical needs (Figure 1).
