Dear Colleagues,

I'm sure there are many biomechanists and others out there who have come
across the problem of finding the single degree of freedom angle between
two segments in 3D space with only three markers when the angle goes
through zero degrees. (For example, the hip angle with markers at the
shoulder, hip and knee.) Using the dot product between the two segments
yields uncertainty about whether the segments have gone past the zero
degree relative position or are coming back again the way they came. This
is the nature of the cos function and can cause "flipping" of the angle
time series data. Has anyone devised a reliable test with real data which
will tell me whether the segments have gone through zero degrees or are
really coming back?