Information for students, faculty and staff regarding COVID-19. (Updated: 1 July 2020)

Skip to content
printicon

Functional data analysis and applications to human movement

Research project The aims of this project are to develop and apply new statistical methods to better utilize the rich information contained in human movement data.

In this research project we develop functional data analysis methods with the aim of analyzing movements in more detail (i.e., as functions), and measure reliability for such functional data. Using such methodology, we can obtain a more extensive interpretation of human movement, and e.g., compare interventions and evaluate the progress of rehabilitation.

Head of project

Lina Schelin
Associate professor, other position
E-mail
Email

Project overview

Project period:

Start date: 2017-01-01

Funding

Swedish Research Council, 2017 - 2020: 6 000 000 SEK

Participating departments and units at Umeå University

Umeå School of Business, Economics and Statistics

Research subject

Statistics

Project description

Movement laboratories with 3D motion analysis system (to capture motion), force plates (to capture forces that cause motion) and electromyography (EMG) equipment (to capture muscle activity patterns during motion), are used both for research and for clinical purposes. The observed data sets are complex and multidimensional, and can typically be described by continuous functions. Still, the continuous functional data series are often reduced into discrete variables (e.g., peak knee flexion angle and time-to-peak force) for statistical comparisons, thereby removing information pertinent to the timing and shape of the functional data. In collaboration with a multinational research team, with high competence in physiotherapy, orthopaedic surgery, sport science, and biomedical engineering, we work towards a more extensive interpretation of human movements.

Within the project we are mainly developing test and regression methods for functional data, with a focus on domain selection; a crucial aspect of inference for functional data related to the area of application. With such method we can for instance identify intervals where covariates have significant effect on a functional outcome variable. The methods we develop can also be applied to similar data from other research fields.