Neuroinformatics Research: 3D Modeling of Neuroanatomy with Multi-Camera System

Academic Background

The surgical anatomy of the central nervous system, including the skull and spine, has an extremely complex three-dimensional (3D) structure, making it difficult for learners to fully understand the intricate relationships between various structures. Accurate understanding of these anatomical relationships and 3D perception is crucial for performing safe neurosurgical procedures and reducing surgical complications. Currently, textbooks and atlases remain the standard methods for teaching human anatomy and neurosurgery (Rhoton, 2023). However, cadaver dissection—considered the most realistic model for studying neurosurgical anatomy (Sotgiu et al., 2020)—is expensive and not universally available in all neurosurgical and anatomy courses. Therefore, there is an urgent need for an economical and effective method of neuroanatomy education for students who cannot access laboratory learning.

Paper Source

This research paper, titled “Photogrammetry Scans for Neuroanatomy Education - A New Multi-Camera System: Technical Note,” is co-authored by André S. B. Oliveira, Luciano C. P. C. Leonel, Megan M. J. Bauman, Alessandro De Bonis, Edward R. Lahood, Stephen Graepel, Michael J. Link, Carlos D. Pinheiro-Neto, Nirusha Lachman, Jonathan M. Morris, and Maria Peris-Celda from the Mayo Clinic in Rochester, Minnesota, USA, Federal University of Paraíba, Brazil, and San Raffaele Scientific Institute in Milan, Italy. The paper was published in Springer Nature on June 6, 2024.

Research Process

Subjects and Materials Preparation

All samples were provided by the Mayo Clinic’s “Body Donation Program” and approved by the Institutional Review Board (IRB 17-005898). The study used a formalin-fixed brain specimen and a dried skull for dissection and scanning.

Experimental Instruments and Equipment

The study employed a new multi-camera photogrammetry system based on five digital single-lens reflex (DSLR) cameras for image capture. Samples were placed in a Medcreator scanner (Medreality, Thyng, Chicago, IL) for 3D model scanning. The new system was configured with a rotating platform to capture images from multiple angles as the specimen rotated.

Photogrammetry Process

The sample was placed on a platform and captured 36 photos per rotation through a 360-degree rotating platform, taking about 15-20 minutes per session. All cameras used Reality Capture software for image capture and processing. The generated 3D models can be displayed on Sketchfab and Medreality platforms for users to freely interact and learn.

Data Processing and Model Display

After model generation, calibration and correction through Reality Capture software eliminated shadows and reflections during the shooting process. The final 3D models were uploaded to Sketchfab and Medreality platforms, allowing users to manipulate these 3D models on any electronic device, enhancing the learning experience.

Main Results

The study found that the new system significantly improved image capture and model accuracy and resolution compared to the previous three-camera system. This not only shortened the model production time but also enhanced the quality of the 3D models. Through these platforms, students and professionals can freely access and manipulate these 3D models on any device, which greatly aids neuroanatomy education and ultimately improves patient surgical outcomes.

Research Conclusions and Value

This paper demonstrates the potential and importance of multi-camera 3D photogrammetry technology in neuroanatomy education. This system excels in improving model accuracy and resolution, shortening model production time, and providing a more convenient and efficient learning tool for education and practice. Additionally, this technology enhances the dissemination and accessibility of neuroanatomy knowledge, cultivating students’ and professionals’ 3D spatial perception, helping to reduce surgical complications, and improving clinical outcomes.

Research Highlights

This study innovatively adopts a multi-camera system with five DSLR cameras, significantly improving the accuracy and resolution of 3D models and simplifying the operation process. This innovation not only provides high-precision anatomical models for practical teaching but also shortens the model production time, increasing the availability and convenience of teaching resources.

Other Important Information

Although this technology has many benefits for education and practice, producing 3D models using this method still requires a high-cost laboratory environment, professional equipment, and technical support. To maximize the educational effectiveness of the models, it is recommended to add labels and annotations to help learners better understand and master complex anatomical structures. Additionally, using virtual reality and augmented reality technology can provide a more immersive learning experience.

Future Prospects

By further optimizing and promoting this technology, it can improve the teaching quality of neuroanatomy and provide efficient, low-cost 3D modeling solutions for other medical fields. In the future, with continuous technological advancements, this system is expected to play an important role in broader medical education and clinical applications.