Anatomical Computerized Exploration to Excise Malignancies in Deep Facial Compartments: An Advanced Virtual Reality...

This study describes the design and application of a novel advanced protocol for virtual three-dimensional anatomical reconstruction of the deep facial compartments, aiming to improve the preoperative understanding and the intraoperative assistance in complex resective surgeries performed for malignant diseases which extend in complex spaces, including the pterygomaxillopalatine fossa, the masticator space, and the infratemporal fossa.

This study is a non-profit, retrospective, and single-institution case series. The authors clearly describe in detail imaging acquisition protocols which are suitable to segment each target, and a multilayer reconstruction technique is presented to simulate anatomical structures, with particular focus on vascular networks. Virtual surgical planning techniques are individually designed for each case to provide the most effective access to the deep facial compartments. Intraoperative guidance systems, including navigation and virtual endoscopy, are presented, and their role is analyzed.

The study included seven patients with malignant disease located in the deep facial compartments requiring radical resection, and all patients underwent successful application of the protocol. All lesions, except one, were subject to macroscopically radical resection. Vascular structures were identified with overall reconstruction rates superior to 90% for major caliber vessels. Prominent landmarks for virtual endoscopy were identified for each case.

Virtual surgical planning and multilayer anatomical reconstruction are valuable methods to implement for surgeries in deep facial compartments, providing the surgeon with improved understanding of the preoperative condition and intraoperative guidance in critical phases for both open and endoscopic phases. Such techniques allow to tailor each surgical access, limiting morbidity to strictly necessary approaches to reach the disease target.