Sensor Accuracy

The first step in testing was to determine how accurate our base readings are with the newly implemented complementary filter. The flexion experiment had a t-stat of -0.408, and a critical two tailed t-stat is 2.45, therefore the flexion readings were not significantly different from the actual angles. The rotation experiment had a t-stat of -1.61, the critical t-stat is the same due to the same sample size. Therefore, the rotation readings were not significantly different from the real angles.

As the abduction increases, both the flexion and rotation angles decrease. This introduces an issue with the device, as a patient abducts a certain degree, the device will read the movement as flexion and rotation which will prematurely trigger the warning system despite there being no danger to the patient.

We decided to accept this risk because abduction and adduction of the leg are uncommon movements to be making for a patient, and we wanted to focus on the two more common and important movements of the leg.