Computer-aided navigation (CAN) is a surgical technology which allows a surgeon touse patient medical image data as a map to guide both surgical planning and execution.It comprises several interconnected processes: visualization of 3D medical image data,tracking of surgical instrumentation, definition of a virtual coordinate system aroundthe patient, and the alignment of the image data to the patient.Despite quantitative benefits, the technology is often not used due to size, cost, andunintuitive visualization of 3D patient data as 2D black and white images. Augmentedreality (AR) devices often integrate requisite hardware for CAN into a compact andmobile head mounted device (HMD) and allow the surgeon to view complex 3D dataas a “Hologram” overlying the patient. This work addresses technical limitations ofsuch low-cost AR hardware with respect to tracking performance and presents evidencesupporting their use in both neurosurgical and orthopedic domains.Initial work focused on quick-response (QR) code tracking using the device{\textquoteright}s front-facing red-green-blue (RGB) color sensor, allowing a stable coordinate system to bedefined based on the 3D pose of a static QR marker in space. This both improved spatiallocalization of AR visualization to a mean perceived spatial drift of 1.4mm in a dynamictracking environment and provided a proof of concept.Building on this work, the development of monocular infrared (IR) tracking for poseestimation of existing surgical instrumentation provided an improved method towardsestablishing a reference coordinate system and a mechanism for precise user input.Compared with earlier videometric tracking, the transition to the device{\textquoteright}s IR sensorprovided a greater tracking field of view (FoV) and more favorable orientation. Thistracking solution was validated in a video converter (Vicon) motion capture lab anddemonstrated a pose estimation error of 0.78 mm ± 0.74 mm and 0.84° ± 0.64°.Following this, phantom trials in navigated external ventricular drain (EVD), and totalshoulder and hip arthroplasty were performed. The results demonstrated a reductionin technique learning curve of the former, and improved outcomes of the latter whencompared to traditional non-navigated techniques. Moreover, AR data registration wasfound to be comparable to modern CAN systems.Clinical trials in both tumor resection planning and EVD were then performed toassess the efficacy of AR-CAN. In the former, AR-CAN demonstrated a reductionin preoperative planning time with superior lesion delineation when compared to cur-rent neuronavigation. Preliminary results in AR navigated EVD placement outcomesdemonstrate 82 % optimal (grade I), 18 % sub-optimal (grade II), and 0 % (grade III).This currently outperforms literature, given single attempt insertion.