Introduction: With the development of information and communication technology, the advanced techniques such as augmented reality (AR) or virtual reality (VR) technologies have been applied in the medical field. AR is a novel technology of overlaying the virtual information in the real world and recently available in the clinical setting. Thanks to the development of the operating microscope, microscope-based AR technology (AR microscope) was used in cranial neurosurgery in 1990s. Implementation of AR microscope support in spine surgery have been reported since 2018. Purpose of this study is to clarify the utility of AR microscope for spinal cord tumor surgery.
Materials and methods: We evaluated 30 patients including 15 male and 15 female who were diagnosed with spinal cord tumor and underwent tumor resection in our hospital. We created three-dimensional (3D) fusion images from preoperative magnetic resonance images (MRIs) and computed tomography (CT) scans using a software (Elements, Brainlab, Munich Germany). After intraoperative surface registration using a navigation platform (Curve, Brainlab, Munich Germany), 3D fusion images were overlaid as AR navigation images in the surgical microscope (KINEVO 900, Carl Zeiss, Oberkochen, Germany). We evaluated the accuracy of registration and the navigation mismatch between tumors and AR navigation images.
Results: Their average age was 61.5 years old (3 to 85 years old). Types of tumor were 5 cervical dumbbell tumors, 12 thoracic spinal cord tumors (5: intramedullary, 5: intradural extramedullary, 2: extradural), and 13 lumbosacral spinal cord tumors (6: dumbbell type, 7: cauda equina tumor). Gross total resection was conducted in 23 cases, and subtotal resection was done in 7 cases including scheduled two-staged surgery, metastatic spine tumor resection, and giant dumbbell type tumor resection. Their Modified McCormick scales revealed no remarkable changes between before and after tumor resection (I: 15->17, II: 6->6, III: 4->6, IV: 5->1). Intraoperative surface registrations were very accurate (less than 1 mm) except 2 cases of thoracic intradural extramedullary tumor which had difference between preoperative and intraoperative spinal alignments. Three cases of mobile cauda equina tumor had the navigation mismatch between intradural tumor and AR navigation image.
Conclusions: AR microscope can visualize the anatomical structures such as muscle, nerve, bone, vessel, and tumor reconstructed by preoperative MRIs or CT images on the surgical field through the heads-up-display. Therefore, we could safely resect the complicated spinal cord tumors. This technology helped us to understand the anatomical landmark even during revision surgery which was in the surrounding scar tissue. AR microscope is now available for these complicated cases. In the future, the application of AR technology in spine surgery is expected to expand, and it must be used appropriately.