How much data am I going to collect?

While there has been research investigating differences between biomechanical models, few studies have investigated the influence of the number of steps that are included in the analysis. This notion is not new, as Zahedi et al. in their 1987 paper in Prosthetic and Orthotics International already concluded that “it is first necessary to quantify the degree of repeatability due to the method of measurement and step to step variation, before attempting biomechanical comparison.” To our knowledge, however, step-to-step variation has not been studied in individuals with an amputation.

Therefore, Roessingh Research and Development conducted pilot experiments in which we asked three individuals with a transfemoral amputation to come to our lab five times while we collected around 200 strides using instrumented gait analysis. This allowed us to compare variability of gait within a measurement session but also between measurement sessions. Results of this pilot study showed that including a limited number of steps (up to 20 steps) can lead to differences up to 10% in walking speed when comparing across sessions. It also showed that during a measurement session, individuals tend to start at a higher walking speed which levels off towards the end of a session. This pilot study gave indications that the gait pattern of individuals with an amputation is more variable than we initially thought. It also suggests that we may need to include more strides than is common practice at this moment, or at least be aware of the potential risks of analyzing limited number of steps. During the symposium we will delve into the results in much more detail and open the floor to discuss experiences of other researchers.

How am I going to present the data?

Another choice that needs to be made is how we are going to present the data. A common way to present the data is using a stride time-normalization. When analysing human walking graphs of the joint kinematics (angles) are typically presented from initial contact to initial contact with an indication where the step-to-swing phase transition is, indicated at the instant of foot-off. While this may work on an individual level, it may lead to problems when presenting data of the group, or when presenting longitudinal data. This is particularly the case when there is variance in where the step-to-swing phase transition is located. This variance may lead to a situation that data of individuals that are still in the stance phase is combined or compared with data of individuals that are already in the swing phase for a part of the stride.

A thorough analysis of data Roessingh Research and Development gathered in a longitudinal study of stroke survivors showed that this does occur and it can seriously skew the graphs of joint kinematics. Furthermore, this trial showed that normalizing joint kinematics on the sub-phases of double limb support phases, single limb support phase and swing phase prevents skewing of graphs giving a much better representation of the joint angle. During the symposium we will show these data so you can see how data are skewed with your own eyes and how sub-phase time normalization alters the graphs giving a better representation.

What data am I going to present?

The final topic of this symposium is the question of which data to present. As mentioned before, many studying comparing different technologies have looked at joint kinematics or joint kinetics. While these are variables of interest, the relation between differences on these variables and the overall functioning of its user is not always straightforward. Therefore, looking at overall variables such as balance or cognitive load during walking might be more interesting.

One way of looking at cognitive load is measuring the blood flow in the frontal cortex of the brain, as measured with functional near-infrared spectroscopy. Jönköping University has extensive experience in collecting these data in individuals with an amputation. Their data showed that the cognitive load of walking with prostheses with advanced capabilities, such as auto-adaptive components, is lower when compared to walking with mechanically passive prostheses. They also demonstrated that patterns of brain activity may change in the absence of changes in walking speed or step length. These results are of high clinical interest as they clearly show the potential of more advanced prosthetic components on the overall functioning of individuals with an amputation, and the potential of cognitive load as an interesting outcome measure in gait research. In this symposium we will show you the magnitude of these differences and we will discuss the meaning of these results for individuals with an amputation and the prosthetics field in general.

Do you want to know more?

Did we spike your interest and are you attending the ISPO World Congress? You can attend our symposium on Tuesday 25th of April from 14.45 – 16.00 hr in the Main room. In case you are not attending, but you are interested? Please feel free to reach out to us!