No movement is executed in isolation. In real life, we produce a constant stream of actions– with the movement being already selected and planned while the current movement is still being executed. Many skills, such as typing, playing an instrument or tying a knot, rely on complex sequences of movements, sometimes practiced and rehearsed, sometimes flexibly assembled just in time.
How does the brain plan and execute movements at the same time? How do we become better at concatenating simple movements into a skillful whole? What neural representations underlie the production and learning of sequence movements? To answer such questions, we are combining carefully designed behavioral experiments, high-resolution functional magnetic resonance imaging (fMRI), representation analysis tools, and computational modelling.
Sequence representation in human cerebral cortex
Using a pattern classification we first established that we could decode the sequence identity from the fine-grained (~2.3mm3) fMRI activity patterns of participants performing several 5-item sequences after a few day of practice (Wiestler and Diedrichsen, 2013). In further studies we then asked how sequences are represented in different brain regions. For example, we have shown that the temporal (i.e., time intervals between presses) and ordinal (i.e., order of fingers in a sequence) features of sequences are stored in partly separated areas including PMd, SMA, and SPL (Kornysheva and Diedrichsen, 2014). We also were able to show that the representation in rostral PMd reflect the instructional cue, while more caudal areas represent the actual finger movement (Wiestler et al., 2014).
More recent works from the lab addressed the question of hierarchical sequence organization. In the mid-twentieth century, Karl Lashley, George Miller, and others proposed that the actions are organized hierarchically by recursively grouping several simpler movements into a higher-order functional unit, sometimes called "chunks". Activation of a chunk automatically generates a set movement elements in specific order. Recently, we designed a new behavioral paradigm to stably introduce a 3-level hierarchy (single finger, chunk, and entire sequence), and to study these levels in brain activity patterns (Yokoi and Diedrichsen, 2019). Our data clear shows that M1 represents only individual finger movements (Yokoi et al., 2018). Within premotor and parietal cortices, chunk and sequence representations partially overlapped.
Work from our and other labs show that sequential behaviours are controlled by a tight interaction between action selection, planning, and motor execution. How does the brain coordinate these processes without catastrophic interference? What type of solutions emerge in artificial neuronal networks trained to perform sequential tasks? How do these representations evolve over long periods of training?
- Ariani, G., Kordjazi, N., Pruszynski, J., Diedrichsen, J. (2021). The Planning Horizon for Movement Sequences. eneuro, 8(2), ENEURO.0085-21.2021.
- Popp, N., Yokoi, A., Gribble, P., Diedrichsen, J. (2020). The effect of instruction on motor skill learning. Journal of Neurophysiology, 124(5), 1449-1457.
- Berlot, E., Popp, N., Diedrichsen, J. (2020). A critical re-evaluation of fMRI signatures of motor sequence learning. eLife, 9.
- Ariani, G., Kwon, Y., Diedrichsen, J. (2020). Repetita iuvant: repetition facilitates online planning of sequential movements. Journal of Neurophysiology, 123(5), 1727-1738.
- Yokoi, A., Diedrichsen, J. (2019). Neural Organization of Hierarchical Motor Sequence Representations in theHuman Neocortex. Neuron.
- Beukma, P., Diedrichsen, J., & Verstynen, T. (2019). Binding during sequence learning does not alter cortical representations of individual actions. Journal of Neuroscience.
- Ariani, G., & Diedrichsen, J. (2019). Sequence learning is driven by improvements in motor planning. Journal of Neurophysiology, 121(6), 2088-2100.
- Yokoi, A., Arbuckle, S. A., & Diedrichsen, J. (2018). The Role of Human Primary Motor Cortex in the Production of Skilled Finger Sequences. J Neurosci, 38(6), 1430-1442.
- Berlot E., Popp N.J., Diedrichsen, J. (2017). In search of the engram, 2017. Current Opinion in Behavioral Sciences..
- Waters, S., Wiestler, T., & Diedrichsen, J. (2017). Cooperation Not Competition: Bihemispheric tDCS and fMRI Show Role for Ipsilateral Hemisphere in Motor Learning. J Neurosci, 37(31), 7500-7512.
- Diedrichsen, J., & Kornysheva, K. (2015). Motor skill learning between selection and execution. Trends Cogn Sci, 19(4), 227-233.
- Kornysheva, K., & Diedrichsen, J. (2014). Spatial and temporal features of motor sequences are stored independently in human premotor areas. eLife
- Wiestler, T., Waters-Metenier, S., & Diedrichsen, J. (2014). Effector-independent motor sequence representations exist in extrinsic and intrinsic reference frames. Journal of Neuroscience.
- Waters-Metenier, S., Husain, M., Wiestler, T., & Diedrichsen, J. (2014) Bihemispheric tDCS enhances effector-independent representations of motor synergy and sequence learning. Journal of Neuroscience.
- Wiestler, T., & Diedrichsen, J. (2013). Skill learning strengthens cortical representations of motor sequences. eLife.