Thomas M. Greiner

11-19-1992, 02:02 AM

I have encountered some unexpected problems with the calculation

of instantaneous centers of rotation. I am hoping that someone

out there might have encountered a similar problem, and

developed a solution.

Just so we all know where I'm coming from, I will outline how I

calculate the instantaneous center of rotation. At the knee,

for example, data is required that describes both the thigh and

the leg before and after knee flexion. These two data sets are

then superimposed so that the thighs match up exactly. Two

landmarks are then identified on the leg from the straight and

flexed knee data sets. Each homologous landmark is joined by a

line and a perpendicular bisector is constructed. The point of

intersection of the two resulting perpendicular bisectors is the

instantaneous center of rotation.

My data consists of three dimensional coordinates for a series

of skeletal landmarks ranging from the sacrum to the foot.

Landmark coordinate data was obtained for each individual in two

postures, one is basically a fully extended limb while the other

is a flexed limb. Using a computer, I am able to precisely

superimpose the coordinates describing any single bone from one

posture over the same bone in the other posture. I assumed that

I could make use of this procedure to examine the displacement

of landmarks associated with an adjacent bone, and thereby

calculate the coordinates for the instantaneous centers of

rotation in the appropriate joint in any given plane of action.

Unfortunately, the answers I get are not consistent and are

frequently extremely inappropriate. Although the joint centers

I have calculated for flexion/extension at the hip and knee

joints seem appropriate, the same methods do not work for the

ankle joint, nor for rotation or abduction at the hip. For one

individual, the calculated center of rotation for the ankle

joint was located near the sacrum. Obviously, this is the wrong

answer, and leads me to question the answers that seem more

appropriate. Even more unexpected, when I use different

different pairs of landmarks I get different locations for the

joint centers.

I have checked, re-checked, and checked again, and I am certain

that there is no error in my coordinate data. Has anyone out

there had similar difficulties? If not, does anyone have any

idea on what I might be doing wrong?

Thomas M. Greiner

BA05158@BINGVMB

of instantaneous centers of rotation. I am hoping that someone

out there might have encountered a similar problem, and

developed a solution.

Just so we all know where I'm coming from, I will outline how I

calculate the instantaneous center of rotation. At the knee,

for example, data is required that describes both the thigh and

the leg before and after knee flexion. These two data sets are

then superimposed so that the thighs match up exactly. Two

landmarks are then identified on the leg from the straight and

flexed knee data sets. Each homologous landmark is joined by a

line and a perpendicular bisector is constructed. The point of

intersection of the two resulting perpendicular bisectors is the

instantaneous center of rotation.

My data consists of three dimensional coordinates for a series

of skeletal landmarks ranging from the sacrum to the foot.

Landmark coordinate data was obtained for each individual in two

postures, one is basically a fully extended limb while the other

is a flexed limb. Using a computer, I am able to precisely

superimpose the coordinates describing any single bone from one

posture over the same bone in the other posture. I assumed that

I could make use of this procedure to examine the displacement

of landmarks associated with an adjacent bone, and thereby

calculate the coordinates for the instantaneous centers of

rotation in the appropriate joint in any given plane of action.

Unfortunately, the answers I get are not consistent and are

frequently extremely inappropriate. Although the joint centers

I have calculated for flexion/extension at the hip and knee

joints seem appropriate, the same methods do not work for the

ankle joint, nor for rotation or abduction at the hip. For one

individual, the calculated center of rotation for the ankle

joint was located near the sacrum. Obviously, this is the wrong

answer, and leads me to question the answers that seem more

appropriate. Even more unexpected, when I use different

different pairs of landmarks I get different locations for the

joint centers.

I have checked, re-checked, and checked again, and I am certain

that there is no error in my coordinate data. Has anyone out

there had similar difficulties? If not, does anyone have any

idea on what I might be doing wrong?

Thomas M. Greiner

BA05158@BINGVMB