JultikaUniversity of Oulu repository

Abstract

This thesis presents a signal analysis tool, which has been designed to investigate walking abnormalities which are related to foot rolling movements during walking; interaction of foot with ground which is called stance phase. They would cause a wide range of severe anatomical damages such as ankle, leg, heel and back pain in the long-term.

Comparing to the conventional data acquisition setups of biomechanical researches, inertial measurement sensors (IMU), which are being used widely as an appropriate alternative setup recently, facilitate monitoring human movement for a long-term period out of laboratory. This justifies the growing trend of improving the IMU-based algorithms which are designed for events detection, position calculation, and rotation estimation. Therefore, a set of 4 IMUs, placed on shank and foot of both legs, has been used for data collection.

In data processing stage, two novel algorithms have been developed and implemented as the backbone of the designed software aiming to detect and integrate stance phases. The first algorithm was developed to detect stance phases in gait cycle data. Even though the detection of events in gait cycles has been the topic of a majority of biomechanical researches, stance phase as the interval between two consecutive events has not been studied sufficiently. The second algorithm, sensor alignment, generates a rotation matrix which is used to align IMU sensors placed on the same foot and shank. This alignment of the two sensors enables us to add or subtract the data point-wisely to make a more meaningful interpretation of the data regarding thought-out walking abnormalities during phase stances.

The visualized results of the thesis can be considered as an early stage of a more comprehensive research which might lead to quantitative results corresponding to different walking abnormalities.