<< Back to the abstract archive
Craniofacial Surgical Planning with Augmented Reality: Accuracy of Linear 3D Cephalometric Measurements on 3D Holograms
Kihyun Cho, MD, MSc (1); Jeff Yanof, PhD (2); Graham Schwarz, MD (1); Karl West, MS (2); Hirsh Shah, BS (1); Maria Madajka, PhD (1); Richard L. Drake, PhD (3); Bahar Bassiri Gharb, MD, PhD (1); Antonio Rampazzo, MD, PhD (1); Francis A. Papay, MD (1)
(1) Plastic Surgery Department, Cleveland Clinic (2) Biomedical Engineering Department, Cleveland Clinic (3) Anatomy Department, Cleveland Clinic
Cleveland Clinic Foundation, Department of Plastic Surgery
2017-02-15
Presenter: Kihyun Cho
Affidavit:
This project is the original work of the research fellow
Director Name: Frank A. Papay
Author Category: Fellow Plastic Surgery
Presentation Category: Basic Science Research
Abstract Category: Craniomaxillofacial
Background: Conventional craniofacial surgical planning software lacks intuitive operator interaction and 3D navigation. Recently available augmented reality (AR) headsets have the potential for detailed 3D holographic projections of virtual 3D models, intuitive interaction and navigation, and accurate measurements.
Objective: To evaluate the accuracy of linear measurements performed on HoloLens (Microsoft, Seattle, Wash.) in comparison with caliper and a 3D cephalometric analysis program (Dolphin Imaging, Chatsworth, Calif.) toward assessing potential use for surgical planning.
Methods: Seven adult dry human skulls were selected based on stable occlusion. Stereolithography files were generated from CT imaging data acquired with a default imaging protocol, loaded onto HoloLens. Two researchers performed three repeated linear measurements on 1) skulls with calipers as an anatomic physical measurement (gold standard), 2) Dolphin Imaging, and 3) holograms by selecting two surface points. A subset of measurements that could be measured with calipers was used. The analysis of variance (P≤.05) and intraclass correlation coefficient was determined as an index of intraobserver and interobserver reliability. The agreement between the calipers versus HoloLens and the Dolphin versus HoloLens were also evaluated using the Bland and Altman analysis.
Results: All HoloLens measurements were completed by easily "gazing" and navigating through complex 3D structures augmented to the virtual operating room. There was no significant difference between the three methods used.
Conclusions: Our results indicate that HoloLens has significant potential for surgical planning including sufficient accuracy for linear measurements. Future studies will include formal usability evaluation and augmentation of 3D holographic planning results for intraoperative use.
