Augmented Reality Body Image (ARBI) - Postmortem

This application gives users the opportunity to compare and adjust a virtual avatar with an actual image of themselves through AR technology. Using real patient data from the EMR, this application presents an avatar of the patient, assuming proportions in relation to the data that is fetched. The user/patient then adjusts the avatar to more accurately represent themselves, while comparing the avatar to a photograph or other real-life representation. The goal of the application is to help those who struggle with body dysmorphia have a more healthy and realistic vision of their body image. Features: Fetching patient data (height and weight) from MyChart; Rough avatar representation generated by application; Customization features for avatar; Comparison scene; User compares avatar to a photograph of themselves in real space; Hand-gesture controls/input: pointing, grabbing, dragging through HoloLens technology.

GENERAL FEEDBACK ● ● ● ● ● More clear direction would have been helpful (i.e. what the app actually is). We felt it was a little too exploratory. We were kind of brainstorming throughout the whole semester due to the nature of the project Testing time is huge. It takes a long time to push the app to the HoloLens. Debugging is difficult when we put a bunch of features in at once. ○ Usually we have to put the features in at once though… Having Sid sitting closer to the team would have been helpful for communication User Feedback: ● Instructions would have been helpful, or better instructions. ● Cursor changing was unclear. Users didn't really know what was going on with that. ● We need to have our own space for demos\ ○ Spatial mapping is screwed up with people walking by. ● People were more interested in the tech than the app. ● Animations would make it more interactive and exciting ● Some sort of additional feedback would have been helpful (sound, animation) ● We didn't get many people through the demo (takes awhile to get it situated, and there's confusion) ● Gestures were super unclear. Even when we did the gestures for them. ● Angle of the HoloLens is awkward, and gestures are sometimes out of view. ● Placing the HoloLens just slightly improperly can totally screw up the experience. Concerns/Things we would change: ● Code changes were made that interfere with staying up to date on Microsoft API ○ If we have a new feature, you can't rely on documentation because the code has been changed! ■ For example, hand pressed gesture code was changed. ● JSON - getting it to work is still a pain. HoloLens isn't playing nice with it. Need to look into this more to see how we can make it work. ● Communication on when build updates are ready could be better (engineers didn't know animations were ready to go).