Rotational dynamics in motor cortex are consistent with a feedback controller

Abstract

SummaryRecent studies hypothesize that motor cortical (MC) dynamics are generated largely through its recurrent connections based on observations that MC activity exhibits rotational structure. However, behavioural and neurophysiological studies suggest that MC behaves like a feedback controller where continuous sensory feedback and interactions with other brain areas contribute substantially to MC processing. We investigated these apparently conflicting theories by building recurrent neural networks that controlled a model arm and received sensory feedback about the limb. Networks were trained to counteract perturbations to the limb and to reach towards spatial targets. Network activities and sensory feedback signals to the network exhibited rotational structure even when the recurrent connections were removed. Furthermore, neural recordings in monkeys performing similar tasks also exhibited rotational structure not only in MC but also in somatosensory cortex. Our results argue that rotational structure may reflect dynamics throughout voluntary motor circuits involved in online control of motor actions.HighlightsNeural networks with sensory feedback generate rotational dynamics during simulated posture and reaching tasksRotational dynamics are observed even without recurrent connections in the networkSimilar dynamics are observed not only in motor cortex, but also in somatosensory cortex of non-huma n primates as well as sensory feedback signalsResults highlight rotational dynamics may reflect internal dynamics, external inputs or any combination of the two.

Learning to Use Muscles

“…The engineering term for this is rotational dynamics. Such activity has been found in primate motor cortex and it is produced by artificial neural networks trained to perform similar tasks; for recent results and review see ( Kalidindi et al, 2020 ).…”

Convergence of proprioceptive and visual feedback on neurons in primary motor cortex

“…Primary motor cortex (M1) plays an important role in generating goal-directed corrections during motor actions. M1 receives rich sensory inputs from many brain regions involved in proprioceptive and visual processing including the parietal and frontal cortices (Jones et al, 1978; Zarzecki and Strick, 1978; Crammond and Kalaska, 1989; Porter and Lemon, 1993; Buneo et al, 2002; Pesaran et al, 2006; McGuire and Sabes, 2011; Bremner and Andersen, 2012; Dea et al, 2016; Omrani et al, 2016; Gamberini et al, 2017; Piserchia et al, 2017; Kalidindi et al, 2020; Takei et al, 2021), as well as input from cerebellum (Conrad et al, 1975; Vilis et al, 1976; Strick, 1983; Guo et al, 2020; Sauerbrei et al, 2020). M1 rapidly responds to proprioceptive feedback of the limb within ∼20-40ms of an applied mechanical load (Evarts and Tanji, 1976; Wolpaw, 1980; Lemon, 1981a; Suminski et al, 2009; Pruszynski et al, 2011, 2014; Omrani et al, 2014; Heming et al, 2019) and to visual feedback of the limb and goal within ∼70ms (Georgopoulos et al, 1983; Cisek and Kalaska, 2005; Ames et al, 2014; Stavisky et al, 2017).…”

Small, correlated changes in synaptic connectivity may facilitate rapid motor learning

“…Our model consistently underestimated the changes in trial-averaged activity observed during VR adaptation, despite closely matching the small covariance changes (Figure 2,4). This is to be expected, as the model only captures changes due to the motor adaptation process itself, whereas the actual neural activity contains signals related to other processes such as impulsivity/engagement [Cowley et al, 2020, Hennig et al, 2021 and feedback processing [Omrani et al, 2016, Stavisky et al, 2017, Kalidindi et al, 2020, Perich et al, 2020, Cross et al, 2021. In fact, the experimentally observed neural activity changes between the early and late trials of control reaching sessions with no perturbation were almost as large as the changes during adaptation in our model (Figure 2C black dots).…”