Coronavirus (Covid-19): Latest updates and information
Skip to main content Skip to navigation

Examining the rules that govern skilled sequential behaviour in humans

Primary Supervisor: Dr Joseph Galea, School of Psychology

Secondary supervisor: Dr Massimiliano Di Luca

PhD project title: Examining the rules that govern skilled sequential behaviour in humans

University of Registration: University of Birmingham

Project outline:

Aim

The goal of this project to understand the rules which govern how everyday life sequential actions (eating, drinking, using a smartphone) are performed with speed, smoothness and energetic-efficiency. The project will involve the assessment of complex behaviour within a laboratory environment, the development of novel technology that examines behaviour within a home environment and the creation of interventions that can provide online feedback regarding optimal performance.

Background

During everyday life, humans perform a remarkable amount of sequential actions with their upperlimbs such as writing, eating, drinking, driving and using a smartphone. When we first encounter such sequential actions, they are performed as a set of slow and inefficient submovements that have a clear stop period between them (Gulde & Hermsdorfer, 2018). For example, when drinking a cup of coffee, we may reach out, grasp the cup, raise it to our lips, tilt the cup and drink from it. However, with learning these discrete elements are gradually blended together so that the overall sequential action is executed with increased speed, smoothness and energetic-efficiency (Hansen, Grimme, Reimann, & Schoner, 2018).

The process of submovements gradually overlapping one another is referred to as coarticulation and is an essential mechanism for understanding skilled sequential performance as it reflects the evolution behaviour towards increased speed and efficiency (Sosnik, Hauptmann, Karni, & Flash, 2004). In addition, impaired coarticulation during sequential actions is apparent in many movement disorders such as Stroke (Rohrer et al., 2004) and Parkinson’s disease (Lange et al., 2006). Despite this, we know very little of the underlying mechanisms and computational rules that govern coarticulation during naturalistic sequential behaviours (Shah, Barto, & Fagg, 2013; Sporn, Chen, & Galea, 2020). This lack of understanding is a significant problem for the development of interventions that can improve coarticulation within a clinical context.

One of the critical reasons underlying this lack of knowledge is the current focus on simple, discrete laboratory-based tasks that involve minimal coarticulation. Therefore, in order to develop a greater understanding of coarticulation, an in-depth analysis of complex sequential activities of daily living is required. By understanding the mechanisms which govern coarticulation during activities of daily living, this project will deliver significant impact across scientific fields that strive to understand and improve human behaviour such as motor control, childhood development, ageing, rehabilitation, robotics and brain-machine interfaces.

Objectives

  1. Examine the rules that govern coarticulation during complex sequential activities of daily living within a laboratory environment.
  2. Develop novel technology/methodology that enables sequential activities of daily living to be examined within the home environment through a smartphone video.
  3. Develop and test interventions that provide online feedback regarding optimal coarticulation during activities of daily living.

Methods

  1. Typical healthy individual’s upperlimb behaviour will be collected during laboratory-based versions of activities of daily living through marker-based motion tracking equipment (Polhemus Liberty).
  2. Kinematic behaviour will also be collected through markerless technology (DeepLabCut/OpenPose) in which deep learning (neural networks) algorithms are used to track upperlimb behaviour during activities of daily living within the home environment.
  3. A range of 2-D and 3-D movement parameters (coarticulation, speed, smoothness) will be extracted and analysed through Matlab/Python.
  4. A coarticulation intervention (enabling online feedback regarding an individual’s performance and optimal coarticulation) will be developed and tested either within the laboratory (Matlab/Python) or at home (Unity/Java).

References:

  1. Gulde, P., & Hermsdorfer, J. (2018). Smoothness Metrics in Complex Movement Tasks. Front Neurol, 9, 615. doi:10.3389/fneur.2018.00615
  2. Hansen, E., Grimme, B., Reimann, H., & Schoner, G. (2018). Anticipatory coarticulation in non-speeded arm movements can be motor-equivalent, carry-over coarticulation always is. Exp Brain Res, 236(5), 1293-1307. doi:10.1007/s00221-018-5215-5
  3. Lange, K. W., Mecklinger, L., Walitza, S., Becker, G., Gerlach, M., Naumann, M., & Tucha, O. (2006). Brain dopamine and kinematics of graphomotor functions. Hum Mov Sci, 25(4-5), 492-509. doi:10.1016/j.humov.2006.05.006
  4. Rohrer, B., Fasoli, S., Krebs, H. I., Volpe, B., Frontera, W. R., Stein, J., & Hogan, N. (2004). Submovements grow larger, fewer, and more blended during stroke recovery. Motor Control, 8(4), 472-483.
  5. Shah, A., Barto, A. G., & Fagg, A. H. (2013). A dual process account of coarticulation in motor skill acquisition. J Mot Behav, 45(6), 531-549. doi:10.1080/00222895.2013.837423
  6. Sosnik, R., Hauptmann, B., Karni, A., & Flash, T. (2004). When practice leads to co-articulation: the evolution of geometrically defined movement primitives. Exp Brain Res, 156(4), 422-438. doi:10.1007/s00221-003-1799-4
  7. Sporn, S., Chen, X., & Galea, J. (2020). Reward-based invigoration of sequential reaching. bioRxiv. doi:https://www.biorxiv.org/content/10.1101/2020.06.15.152876v1

BBSRC Strategic Research Priority: Understanding the Rules of Life: Neuroscience and behaviour

    Techniques that will be undertaken during the project:

    • Motion tracking during laboratory tasks with marker-based technology (Polhemus).
    • Motion tracking during complex activities of daily living with markerless smartphone-based technology (DeepLabCut).
    • Analysis of kinematic features of upperlimb behaviour (Matlab/Python).
    • Computational modelling and recognition of limb kinematics through deep learning (neural networks).

    Contact: Dr Joseph Galea, University of Birmingham