Augmented-Reality Surgery to Guide Head and Neck Cancer Re-resection: A Feasibility and Accuracy Study

Abstract

Background: Given the complex three-dimensional (3D) anatomy of head and neck cancer specimens, head and neck surgeons often have difficulty relocating the site of an initial positive margin to perform re-resection. This cadaveric study aimed to determine the feasibility and accuracy of augmented reality surgery to guide head and neck cancer re-resections.

Methods: This study investigated three cadaveric specimens. The head and neck resection specimen was 3D scanned and exported to the HoloLens augmented reality environment. The surgeon manually aligned the 3D specimen hologram into the resection bed. Accuracy of manual alignment and time intervals throughout the protocol were recorded.

Results: The 20 head and neck cancer resections performed in this study included 13 cutaneous and 7 oral cavity resections. The mean relocation error was 4 mm (range, 1-15 mm) with a standard deviation of 3.9 mm. The mean overall protocol time, from the start of 3D scanning to alignment into the resection bed, was 25.3 ± 8.9 min (range, 13.2-43.2 min). Relocation error did not differ significantly when stratified by greatest dimension of the specimen. The mean relocation error of complex oral cavity composite specimens (maxillectomy and mandibulectomy) differed significantly from that of all the other specimen types (10.7 vs 2.8; p < 0.01).

Conclusions: This cadaveric study demonstrated the feasibility and accuracy of augmented reality to guide re-resection of initial positive margins in head and neck cancer surgery.

FIG. 1

Accuracy study workflow

FIG. 2

Microsoft HoloLens 2 mixed-reality headset. The surgeon is able to maintain sterility in the operating room due to the hands-free nature of the HoloLens

FIG. 3

Video of alignment of three-dimensional (3D) specimen hologram. The surgeon manually aligns the 3D specimen hologram back into the resection bed defect

FIG. 4

Relocation error as a function of specimen type (maxillectomy/mandibulectomy vs other). A Mann-Whitney non-parametric test was performed with an α value of 0.05

FIG. 5

Glossectomy three-dimensional (3D) specimen hologram illustrating difficulty of operating without a headlight