Science Spotlight: 3D Atlas of Neurological Surgery

Introduction

Hi! My name is Toma Spiriev and I am a neurosurgeon from Sofia, Bulgaria, currently practicing in the Department of Neurosurgery, Acibadem CityClinic University Hospital Tokuda. I also teach human anatomy at the Department of Anatomy, Medical Faculty, Sofia University. My primary research is focused on implementation of 3D technologies (3D modeling, preoperative planning visualization in neurosurgery and neuroanatomy) in our everyday practice. My PhD thesis was on preoperative planning in cranial neurosurgery using open-source software packages such as Horos (Horos ProjectTM), OsiriX (Pixmeo Bernex, open-source at that time), 3D Slicer (Slicer Community) to correctly locate and simulate the steps of the surgery such as patient positioning, skin incision, opening of the skull, location of the pathological process and closure steps. This is where I got my first 3D modeling skills and where I started to study Blender software (Blender Foundation, Amsterdam, Holland).

Figure 1. Example of preoperative planning in brain tumor surgery on Horos Software (Horos ProjectTM) based on contrast enhanced preoperative magnetic resonance imaging– A) localization of the projection pathological process over the scalp marked with green dots. B) Skull stripping technique leaving the brain exposed and superficial venous vessels. C) Simulation of the skin incision and craniotomy

Figure 2. Example of preoperative planning in skull base surgery (meningioma of the anterior cranial fossa) on Horos Software (Horos ProjectTM) based on Computed tomography angiography (CTA). A) CTA sagittal section, presenting the tumor and vessels displacement. B) Preoperative planning and simulation in 3D showing craniotomy, tumor and the planning of the vascularized tissues (pericranium) length needed for closure (blue line). C) and D) Tumor seen from inside the skull (posteriorly) and frontal view with craniotomy simulation.

I am involved in the development of a neurosurgical 3D collection database facilitating learning of complex neurosurgical anatomy and approaches presented in the 3D Atlas of Neurological Surgery.

This is and open-access project and represents a collaboration between several European neurosurgical and anatomical departments including:

•     Department of Neurosurgery, Acibadem CityClinic University Hospital Tokuda, Sofia, Bulgaria
•     Department of Anatomy and Histology, Pathology and Forensic medicine, University Hospital Lozenec, Medical Faculty, Sofia University, Sofia, Bulgaria,
•     Department of Anatomy, Heinrich Heine University of Düsseldorf, Germany
•     Department of Neurosurgery Heinrich Heine University of Düsseldorf, Germany
•     Department of Cellular and Molecular Medicine University of Copenhagen
•     Department of Neurosurgery, Rigshospitalet, University of Copenhagen Denmark
•     Laboratory of Neuroanatomy, Ebris institute of Salerno, Italy

The project is nonprofit and is supported by the institutions involved, the European Association of Neurosurgical Societies (EANS) 2022 Research Fund and Sketchfab educational program.

The project is unique because it contains a large database of photorealistic layered 3D models of complex neuroanatomy.

Workflow

The study of neuroanatomy has always been a challenge to every medical professional. The main problem is that in order to understand certain anatomical areas, one has to do a mental 3D reconstruction by studying numerous visual perspectives of regions of interest from different sources – books, 2D atlasses. However, this might be a very difficult and timely process. The advances of technologies can facilitate immensely this process by several means. First, the availability of open-source software for 3D medical visualization such as Horos (Horos ProjectTM), 3D Slicer (Slicer Community) can facilitate 3D visualization by the possibility to segment medical images on the basis of CT or MRI data. Second, the recent giant leap in the open-source Blender 3D graphics software development, making it more user friendly and the availability of large freely accessible tutorial databases on the internet can further facilitate the post processing of medical 3D models even with moderate 3D modeling skills. And third, the advancement of the internet, allowing online 3D models visualization and interaction, as well making augmented and virtual reality more and more accessible to a general audience.

Therefore, our journey using the Sketchfab platform started a year and a half ago, when we first started to upload 3D models based on routing CT or MRI studies, which is sufficient to present some basic anatomical concepts (brain vessels, bone anatomy) and simulation of some basic approaches in neurosurgery:

Our algoritm of how we create these 3D models using open source software Horos (Horos ProjectTM) is described in detail in the following papers:

• Preoperative 3D planning in cranial neurosurgery
• Three-Dimensional Printing in Neurosurgery: A Review of Current Indications and Applications and a Basic Methodology for Creating a Three-Dimensional Printed Model for the Neurosurgical Practice

Later, after some research we started to implement photogrammetry algorithms in the Department of Anatomy Hospital Lozenec, Medical Faculty, Sofia University, Sofia, Bulgaria, to generate unique photorealistic anatomical 3D models. We started with some basic anatomical preparations such as bone structures and later with some macroscopic dissections.

We were encouraged with the initial results and we later applied for funding, together with the Department of Neurosurgery Heinrich Heine University of Düsseldorf, Germany to the European Association of Neurosurgical Societies research fund where our project was selected as one of the winners. This allowed us to create an open-access website where all the 3D models database is displayed, powered by the great functionality of the Sketchfab platform.

For the creation of specialized photorealistic dissections, we used a photogrammetry algorithm and the data was further processed with Blender software.

In Bulgaria, at the Department of Anatomy, Medical Faculty, Sofia University some general anatomical dissections were done including layered back muscle anatomy (which was described in an open access paper), neck anatomy and bone anatomy.

Later at the Department of Anatomy, Heinrich Heine University of Düsseldorf, Germany and Department of Neurosurgery Heinrich Heine University of Düsseldorf, Germany, specialized dissections were done presenting the anatomy of the vertebral artery and complex approaches to the cranio-cervical junction:

Figure 5. Posterior neck region dissection exposing the deepest muscle layer around the vertebral artery.

All dissections were photogrammetrically scanned and present this difficult-to-study anatomy of the craniocervical region. For this study, we used a simplified photogrammetry method using the smartphone camera of an iPhone 11 only. However, the level of detail of the 3D models is very high and the possibility to pan, zoom, and rotate the 3D anatomical model from every possible angle creates a unique learning experience.

The results of the project are already published in the Operative Neurosurgery journal.

Later in the year the project continued in the Department of Cellular and Molecular Medicine University of Copenhagen and the Department of Neurosurgery, Rigshospitalet, University of Copenhagen Denmark, where another set of specialized anatomical dissections were made in the most difficult part of neurosurgery, mainly skull base surgery and neuroanatomy:

Figure 6. Skull base approaches database.

With the permission of the Department of Cellular and Molecular Medicine, we scanned some of the skull bone collection which are very informative for studying skull bone anatomy:

Figure 7. Sphenoid bone. Example of individual bone photogrammetry scans.

At that time dissections were used during the annual Copenhagen Skull Base Course and Copenhagen Peripheral Nerve course, where participants could observe 3D models on a large screen and follow every step on the dissections.

Figure 8. Annual Copenhagen Skull base course, September 2022, which is held in the Panum Institute, Copenhagen University. An annotated 3D model of the petrous bone is presented on a large screen during the dissections of this anatomical area. Lektor – Lars Poulsgaard

The results of this project are presented in the following paper in Operative Neurosurgery journal.

The project was awarded in October 2022 at the annual EANS conference at Belgrade, where the project’s website was officially launched for the first time.

Figure 9. Presentation of the project and the platform on the EANS Research fund awards, Belgrade, EANS Congress 2022.

At present, all the 3D databases are actively used to teach medical students and neurosurgery residents with the 3D atlas platform and in VR environment.

The EANS Research funding from the project allowed us to create a small 3D lab at the Department of Neurosurgery, Acibadem CityClinic University Hospital Tokuda, Sofia, Bulgaria, where we have all the equipment (3D printers, VR headsets, powerful computer configuration) to even further develop our 3D anatomical collection, and preoperatively plan our cases and simulate the surgery, now in VR settings, which was not possible some years ago.

Figure 11. 3D lab at the Department of Neurosurgery, Acibadem CityClinic University Hospital Tokuda, Sofia, Bulgaria

This project was possible also due to the extensive work that all the contributors and the involved departments did:

3D Atlas of Neurological Surgery – Contributors

3D Atlas of Neurological Surgery – Institutions

and the great collaboration with the Sketchfab platform.