Wednesday, September 26, 2018 7:35 AM–9:00 AM ePosters

BACKGROUND CONTEXT Clinicians of all disciplines learn to visualize internal anatomy to enhance diagnostic accuracy and reduce iatrogenic injury. Unfortunately, the performance of this skill is poor. To address this issue, our team developed technology that superimposes diagnostic images on to a patient's skin while minimizing distortion caused by patient topography. The result is that clinicians can immediately visualize their patient's underlying anatomy while freeing their hands to conduct clinical procedures such as palpation, injections or even surgery. PURPOSE In this project, we assess the accuracy and reliability of a holographic projection system with respect to representing underlying anatomy. STUDY DESIGN/SETTING Experimental system compared to criterion reference. PATIENT SAMPLE Ten (10) males (76.9%, mean age 42, age range 31–58) and 3 females (23.1%, mean age 41.6, age range 27–68) were enrolled in the study. OUTCOME MEASURES The projected location of any spinous process was judged to be accurate or “on target” if its location was within known spinous dimensions for that specific level. METHODS Thirteen participants were enrolled in the study, each having a pre-existing anteroposterior lumbar radiograph. Each participant's image was uploaded into a goggle system which when worn, allowed the operator to view the radiograph superimposed on the participant's back. The projected image was topographically corrected using depth information obtained by the goggles then aligned via existing anatomic landmarks. Using this superimposed image, vertebral levels were identified and validated against spinous process locations obtained by ultrasound. This process was repeated 1-5days later. The projection of each vertebra was deemed to be “on-target” if it fell within the known dimensions of the spinous process for that specific vertebral level. RESULTS The projection system created on-target projections with respect to individual vertebral levels 74% of the time with no significant difference seen between testing sessions. The average agreement for all vertebral levels between testing sessions was 75%. CONCLUSIONS The projection system tested here shows promise as a unique technology to assist clinicians in locating internal anatomy. While there are presently many sources of error including those from the device (resolution), viewer (inaccurate marking and placement) X-ray (magnification) and patient (when using generic images), we anticipate these errors can be mitigated, if not corrected, in future studies to provide an accurate and reliable way for clinicians to visualize patient-specific anatomy while allowing their hands to move freely.