Analysis of time-dependent pressure distributions during a footstep
PR/DA
Martin Haidacher, Meister Eduard Gröller
Content:
Task
Investigation of the correlation between high-dimensional data sets with the same indication from different patients. The data sets for this investigation are coming from the measurement of time-dependent pressure distributions during a footstep.Description
Walking is a very complex motor activity which includes many bones and muscles. Slight variations in this motor activity can cause damages to joints and other parts of the legs. These damages are the source of pains. To prevent these damages it is necessary to detect deviations in the walk. A physical therapy or inlays for shoes can then be prescribed to prevent further damages to the body.To detect the deviations in the walk it is a common practice to let patients walk over a so called pressure plate. The pressure plate is able to record the dynamic pressure distribution under the foot for a single step. An expert is using this information for a diagnosis.
This practice of using a pressure plate to detect problems in the walk is in use since quite some while. But until now nobody ever investigated the huge amount of data which was already recorded to find out if there are any correlations between different patients.
In this project it would be your task to investigate this huge amount of data the first time. The goal of your work should be to find out whether there is a correlation between different patients with the same indications or not.
For this investigation you would need to try some data mining techniques (such as neural networks) and also use some programs which are designed for the investigation of high-dimensional data sets.
