Interest in the use of optical imaging instruments in medical interventions stems from their ease of use, rapid adaptation to clinical needs, portability, real-time feedback, and relatively low cost. Of particular interest is the role of optical imaging in oncology. Surgery is the primary curative method for solid tumors confined to the tissue of origin with the goal of completely removing both the tumor mass and microscopic lesions. Unfortunately, the irregular growth pattern and infiltrations into surrounding healthy tissue prevent complete removal in many cases, resulting in positive surgical margins (PSMs). PSMs are prevalent in oncologic surgery, increasing cancer recurrence rates and often necessitates a second surgery to improve disease-specific survival. While PSM occurrence is significant in advanced clinical centers, the situation is worse in many rural hospitals and resource- limited areas due to limited histology infrastructure and workforce needed for margin assessment. Thus there is an urgent need for an intraoperative imaging system to visualize cancer, guide tumor removal, and determine margin positivity in the operating room (OR) in low and high resource settings alike. Handheld fluorescence imaging systems have been developed to aid cancer resection. Still, they suffer from several limitations, including a significant footprint in the OR and the inability of the operating surgeon to directly control the imaging device while performing surgery. To address these shortcomings, we developed a head- mounted display device (HMD) cancer imaging system for real-time intraoperative fluorescence-guided surgery (FGS). The system HMD captures near-infrared (NIR) fluorescence and color images from the surgical bed and displays accurately aligned color-NIR images in real-time, enabling FGS without disrupting surgical workflow. The HMD has a small footprint, is intuitive to use, and is amenable for widespread use, including non-cancer applications such as imaging of peripheral blood flow. Preliminary testing of the HMD system in human cancer patients identified some areas for improvement that will accelerate the eventual clinical adoption of the system worldwide. Addressing these needs requires expertise in packaging software development for medical devices with DICOM image format and user interface development using human factors engineering. We have teamed up with a company that has both expertise and experience in developing augmented reality/mixed reality (AR/VR) software combined with deep machine learning in wearable devices on this project. Together, we will optimize the system performance and ergonomics using human factors engineering. The collaborative project will (1) develop and validate an automated fluorescence thresholding algorithm for tumor delineation; (2) develop and validate automated registration of augmented reality in the system; and (3) develop and evaluate clinical software to improve user experience. At the completion of this project, we expect to develop and validate a clinic-ready, user-friendly HMD system with a small hardware footprint, enabling seamless integration with surgical workflow to enhance clinical adoption. The system will increase the rates and decrease the time of successful tumor resection. Anticipated low cost and ease of use will expand adoption in low and high resource settings worldwide. This objective approach to cancer surgery will reduce the incidence of PSMs and improve treatment outcomes.

Narrative Optical imaging instruments have permeated medical practice because of their ease of use, rapid adaptation to clinical needs, portability, real-time feedback, and relatively low cost. We have developed a wearable head- mounted display device that is augmented by visible and near-infrared fluorescence imaging of tissues in real- time. In collaboration with an industry partner, we will optimize the device performance and ergonomics using human factors engineering. Successful completion of the research will accelerate its clinical translation and adoption in both high- and low-resource communities.