Augmented Reality Systems for Improved Operating Room Workflow

Abstract

INTRODUCTION Minimally invasive surgery (MIS) is considered superior to open surgery because of the reduced recovery times, pain medication, and infection rates. However, the smaller surgical site increases the surgical difficulty as physical landmarks are no longer visible. Using an augmented reality (AR) overlay where a patient's CT or MRI information is registered to the physical patient, users would be able to view the internal patient anatomy while maintaining a small surgical site. Our goal is to determine how to best render the virtual information intuitively for surgeons while operating MIS cases. METHODS Given AR systems available today, there are many options regarding rendering virtual objects. Using the Microsoft HoloLens, we virtually overlay two layers of the patient body (soft and hard tissue) and vary the opacity and color in real time. Surgeons wore the headset during nonoperative stages of the surgery and asked usability questions regarding the system. RESULTS Users surveyed displayed strong predilection towards more opaque renderings of hard tissue, with an average setting of approximately 95%. Comparatively, users preferred much lower opacity for soft tissue, with an average opacity of 26.88%, and with some users wanting no visualization of soft tissue. It was also found that users at earlier stages in their training (ie, medical students and residents) preferred more contrasting colors, whereas attendings and fellows preferred grey-scaled color profiles analogous to traditional imaging modalities. CONCLUSION Based on the results found in the usability testing, it was found that current state-of-the-art AR headsets are not currently well suited for the operating room. Ergonomics, eye fatigue, and changing lighting conditions were all undesirable factors, especially for extended use during surgery. Data gathered from the rendering tests and pain points will be further addressed in the next generation of AR headsets.

Marker-free surgical navigation of rod bending using a stereo neural network and augmented reality in spinal fusion

“…Salah et al (2011) presented an in-vitro study in which imagereconstructed 3D models of the spine including vertebrae, disc, nerves, and the spinal cord were in-situ superimposed on a synthetic torso. Similar proof-of-concept studies have been described for minimally invasive spine surgeries ( Nguyen et al, 2019;Deib et al, 2018 ) as well as for hip ( Chen et al, 2015 ) and orthopedic tumor ( Cho et al, 2018 ) surgery. Recently, AR has been proposed in the context of applications with higher demands on accuracy such as surgical navigation.…”