Since about half the exercises were impossible to do in front of a table given the required range of motion, half way through the application the patient would need to move in front of a table and then calibrate for it height. This was easily done by placing their hand on the table top and setting the y value to that point, thought it was interruptive to the flow of the exercises. It was deemed necessary after watching the patient perform an RULM and seeing how much he relied on the table for support. Luckily, the calibration was very good and opens the door to other integrated reality options in the future.
Many exercises required the use of the table, either as a brace, or as a place for props to be placed. Exercises B, C, D, E, J, K, L, and M were all assessed in the presence of a table. Exercises B and J were lumped together as B required moving their hands from their lap to the table top and J required moving from the table top to mouth height (approximated by half way between shoulder and eye height). This made one continuous motion and saved on development time.
Exercises K, L, and M were also combined as they all required moving an object across the table in different direction. K was approximated L as the motion was the same only with different weights. All of these tests had a deduction for the patient dragging the object. This was assessed as the object intersecting with the table top for more than 3 seconds total or more than 3 instances.
Exercise D was a complicated exercises to assess. In order to grab both coins, 2 different "grab" criteria had to be met. For the first one, the patient had to bend their index finger and apply force on the thumb sensor (pinch grab) and for the other they had to bend their pinky, ring, and middle finger while apply force on the palm sensor (palm grab). The application was able to assess both of these simultaneously though the patient was not able to perform them together either due to their inability of some issue with the G3 or it's communication with the app.
Exercises E was a strictly digitized version of it's RULM equivalent. The patient was presented with 2 cups and 2 coins and instructed to put a coin in each up. They received full credit if they placed a coin in the upper cup and half credit for a coin in the lower cup.
Exercise C was the most ambitious exercises. According to interviews with physical therapists, it has nearly a 100% pass rate so we figured we could do more with it to demonstrate the power VR provides. This made it much harder, likely too hard. Instead of using their hand to trace a path on a sheet of paper, the patient was presented with a laser pointer and a path of good targets (green) surrounded by bad targets (red). They needed to use the laser to break all the good targets to pass the exercise. The patient would not be deducted for breaking bad targets but this could be monitored. We also were able to make the path larger, requiring cross body movement, something not assessed in the RULM. Finally, we would be able to measure the time to completion. This exercise, though likely too hard to be of real value, provides the ability to measure the patients ability in a much more precise and quantitative way than the traditional RULM. In the end we kept the scoring the same but this scene serves as a demonstration of what a custom designed exercises set may include.
Exercise E
Exercise L/M
Exercise C