Clinical Gait Assessment Comparison: Smartphone-Based Versus Inertial Measurements Units

A rising number of people in the United States have physical difficulties related to gait motion. Current clinical gait assessments are subjective and restrictive, as they depend on the presence of a trained physical therapist and technology that is not easily accessible. In order to solve these issues, several low cost, easily transportable systems based on Inertial Measurement Units (IMUs) and smartphones have been proposed in the literature. However, it is unclear which one provides better results. This paper analyzes and compares the performance of a gait analysis system using smartphone sensors and IMUs. The analysis method also compares results when using one and five IMUs. Results show a 1.4779 root-mean-square error for step detection and a better defined and classified step segmentation. However, the highest accuracy for step deviation was achieved by the system when using only one IMU.