The mouse worked very well. Overall, we were surprised with how well we could control the mouse pointers. Due to our painstakingly specific constants to fine tune the movement of the mouse, the mouse itself is very easy to use. It requires very little movement from the resting state of the hand, so it doe not feel awkward to use, but it is not so sensitive that it takes hours of practice to comfortably move the pointer.

The final design does have relatively long wires connecting the accelerometers to the PCB board, but the noise from 60Hz and other sources are low enough that we do not have much trouble.

We were able to get a $6 PS/2 to USB converter that, by some stroke of luck, fully converts our PS/2 mouse into a USB mouse. Furthermore, we found, after testing, that our mouse is fully compatible with mac's and PC's. The USB cable allows for hot swapability - you do not have to reset the computer for the bios to read the PS/2's device ID (which you need to do if it is connected to the PS/2 port).

We let some other people try the mouse on and they were able to use it quite well once they got used to the sensitivity of the mouse. Most people are used to physically moving their hand over a large amount of space to move a normal mouse, but our mouse removes that strain from the hand and the tilting motion feels more fluid and natural to many users. We did encounter some problems with positioning the buttons on some people's fingers. Although the position accelerometers were mounted in such a way to properly move the mouse, the attachment of the button accelerometers using Velcro strips allowed for some rotational movement of the accelerometers. This lead to difficulty with pushing the mouse buttons. Tweaking the fit a little bit for different users usually got around this problem.

As stated previously, the product was originally connected to the hand using simple elastic bands. However, these quickly cut off circulation, so we moved toward using velcro straps that had previously been used for computer cable ties. Another safety issue we had was dealing with the cabling attached to the acceleromters. The user has to be careful not to pull these wires while moving the mouse, although we did make the cabling quite long. Even though there is no on-board power supply, the electrical tape sealing we have used would not be viable for commercial processing. Overall, we believe that with commercial processing, our mouse actually increases the safety of the public in terms of stress and strain reduction on the wrist joint and nerves.

Since we did not have time to add wireless RF communication, we did not have to worry about interference with other devices. We twisted the wires leading to our accelerometers to decrease inductive coupling, and have even considered placing aluminum foil around the twisted wires to create a co-axial cable type of configuration. CPU noise was not a factor in our design, however one must be aware of ensuring that the PS/2 pins are connected correctly to the CPU.