Stair ascend-descend, ground reaction forces, and knee adduction moment

[Daniel Cury Ribeiro]

Greetings from a sunny Dunedin - New Zealand!

I am currently working on a project with some colleagues for exploring knee adduction moments in subjects who have undergone ACL reconstruction, during stair ascend-descend.
When looking at the literature, I have noticed that some studies have used embedded force plates on stairs, while others have placed the stairs over the force plate.
Our current plan is to place the stair over the force plate (ensuring there is no gap between the two), so we can estimate knee adduction moment when individuals step on the last/first stair-step.
This is the first time I am involved with a project with such design, and I wonder if anyone would have some serious concerns with regards to this approach.
Our main outcome measure for this study is knee adduction moments.

Thank you in advance for reading this thread.

Have a great week!

Cheers,

Dan Ribeiro, PhD
Lecturer, School of Physiotherapy - University of Otago

[Maryam Hajizadeh]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Dear Danniel,

I did similar measurements during stair negotiation last year. These are some issues that should be considered in my mind:

(1)The step which is placed over the force plate should be isolated from other steps. Otherwise, the force plate data would not be accurate due to the effect of movement on other steps on the force plate.

(2) The step on the force plate should be heavy enough in order to prevent moving the step relative to force plate during measurements.

(3) The step should be small enough to fit into the force plate borders

(4) The force plate should be zeroed after putting the step on it.

(5) I am not sure what is the best way to have an exact center of pressure during such these measurements. But the easiest solution seems to have a symmetrical mechanical structure of the step, and put it in the most symmetrical position relative to the borders or corners of force plate.


Best Regards,
Maryam Hajizadeh

[Pieter Fiers]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

AMTI has an instrumented stair with an interesting design:

http://www.zflomotion.com/blog/bid/330941/AMTI-Force-Plate-Mounted-Stairs-and-Instrumented-Walkers

Best regards

[Daniel Cury Ribeiro]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Dear Maryam,

Thank you for your time, and advice.
Really appreciate it.

All the best,

Dan

[Bob Cargill]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Dan,
One thing you might need to consider is the effect of the added mass of the step on the frequency response of the system. The mass of the step that you add will effectively act to decrease the frequency response of the system. I looked at some specifications for force plates and the natural frequency was in the 500 Hz range. If your step is 10 times the mass of the force plate sprung mass (just the mass of the plate, not including the frame beneath) then your frequency response will be in the 50 Hz range. That probably won't matter if your subjects are moving slowly, but it will give you erroneously low forces for fast-moving subjects or those who stumble. The rule of thumb is to make whatever you attach as light as possible to minimize this effect. If you can, physically attach your step to the force plate and make it as light at possible.
Best regards,
Bob

[Ton van den Bogert]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Bob makes a very good point.

Natural frequency goes with the square root of the mass, so it's more like a factor 3 in this example. The attachment between the step and the force plate may have less than infinite stiffness, which would lower the natural frequency. Natural frequency in Hz is 2*pi*sqrt(k/m) where k is the stiffness in N/m and m is the mass in kg.

So light and stiff is the goal. If the natural frequency remains above 10 Hz, the setup will still be OK for inverse dynamic analysis. Just make sure to set the low-pass filter below that frequency. I suspect that it will not be a problem to put a step on top of a force plate, if done with care.

Ton van den Bogert

[Daniel Cury Ribeiro]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Thanks Bob and Ton,

I really appreciate your feedback.

Best regards from a sunny Dunedin - NZ.

Dan

[Mohammad S. Shourijeh]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Very nice point regarding the changes in the system natural frequency! Just a small correction that natural frequency of a simple mass-spring system in Hz is (sqrt(k/m))/(2*pi).

Good Luck!
Mohammad S. Shourijeh

Bob makes a very good point.

Natural frequency goes with the square root of the mass, so it's more like a factor 3 in this example. The attachment between the step and the force plate may have less than infinite stiffness, which would lower the natural frequency. Natural frequency in Hz is 2*pi*sqrt(k/m) where k is the stiffness in N/m and m is the mass in kg.

So light and stiff is the goal. If the natural frequency remains above 10 Hz, the setup will still be OK for inverse dynamic analysis. Just make sure to set the low-pass filter below that frequency. I suspect that it will not be a problem to put a step on top of a force plate, if done with care.

Ton van den Bogert

[Dr. H Rassoulian]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Greetings to everyone and to you too Daniel,

Your message poses an interesting problem with which I have some issues but I do have a suggestion too.

First the issue is the idea of measuring the kinetics during stair ascend/descend but only during the first step (if going up) and/or the last step (if going down). If we are to accept these two steps as representing the stair ascend/descend then we should also accept the first and last steps in a normal level walk as representative of the gait. However, we seem not to do that and colleagues sometimes go to an extraordinary length in order to make sure that the forceplate data is captured at some step in the middle of a normal level walk. Should this not be the case with the stair ascend/descend. Though it may well be possible to show that the very first or the last steps while negotiating the stairs are the worst or best case scenarios (in terms of the kinetics) and that might be what is required. Since that has not been shown yet (to my rusty knowledge) then I wonder about the value of such measurements.

Now a suggested solution which is technically abstract but the concept is sound and an engineer working to these abstract specifications should be able to produce a system for you.

Take a stairway that is suitable for your use and construct one single step that is exactly the same dimensions as the first step of the stairway, taking care to make it as light as possible but as rigid as possible (I suggest using carbon fibre reinforced rods which can be purchased easily and then adapted).

Modify the stairway by removing its first step.

Take these two (i.e. the modified stairway and the single step) to the forceplate and bolt the single step to the forceplate. Sometimes you have access to bolt holes on the forceplate (even if you do not have these you can clamp onto the forceplate but you need to open the access around it).

Position the stairway in front of the step such that the single step blends into the sequence of the steps in the stairway as its first step.

Bolt the stairway into the floor.

That should give you a working solution.

With best wishes
Hamid Rassoulian

[Ajit Chaudhari]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Bob makes a very good point.

Natural frequency goes with the square root of the mass, so it's more like a factor 3 in this example. The attachment between the step and the force plate may have less than infinite stiffness, which would lower the natural frequency. Natural frequency in Hz is 2*pi*sqrt(k/m) where k is the stiffness in N/m and m is the mass in kg.

So light and stiff is the goal. If the natural frequency remains above 10 Hz, the setup will still be OK for inverse dynamic analysis. Just make sure to set the low-pass filter below that frequency. I suspect that it will not be a problem to put a step on top of a force plate, if done with care.

Ton van den Bogert

I can attest that experimentally we get results in line with what Bob and Ton have said. With our structure bolted to the force plate, we can see that the natural frequency has dropped lower, but it is above the frequencies we are interested in for walking up and down the stairs. Since we were filtering the motion data at 6Hz, we filtered the GRF data at the same frequency to avoid the possibility of any artifacts from different filtering frequencies (after Kristianslund, Krosshaug, van den Bogert J Biomech 2012). However, as an aside we saw no artifacts when we played around with filtering the GRF at 12Hz instead.

Here is our first published paper using this staircase, which has a nice Figure showing our staircase. We built it mostly out of plywood and pine 2x4s and 4x4s.

Thoma, Louise M., et al. "Quadriceps femoris strength and sagittal-plane knee biomechanics during stair ascent in individuals with articular cartilage defects in the knee." Journal of sport rehabilitation 23.3 (2014): 259-269.


Cheers,
Ajit Chaudhari
The Ohio State University
Host of the ASB 39th Annual Meeting, August 5-8, 2015

[Ton van den Bogert]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Mohammad,

Thanks for your correction!

Ton

[Ton van den Bogert]

Re: Stair ascend-descend, ground reaction forces, and knee adduction moment

Ajit, I am curious If you determined what the natural frequency of your system was. Did you hit it with a rubber hammer and measure the force plate response? If the natural frequency was far higher than needed, this means that you could even get away with a heavier construction.

Ton