It is known that muscle spindles provide the majority of information about limb position, but little is known about how position sense is computed from their signals. We have developed a family of musculoskeletal models in order to determine some of the fundamental properties associated with transforming noisy spindle information into putative internal coordinate frames for position sense. A two-joint model was developed containing one biarticular and two monoarticular muscles with a total of 1000 sensors distributed among them. The sensors were assumed to function like spindle secondary afferents under fusimotor control designed to optimize their ability to encode static position in the presence of constant output noise. The optimal distribution of sensors was found to depend strongly on the coordinate frame in which position was measured (intersegmental angle, segment orientation, or end-point of the limb) and on the topology of the biarticular muscle with respect to the plane of motion. A similar analysis was performed for an anthropometric model of the human arm, using previously published counts of muscle spindles. In general, the actual distribution of spindles about the elbow and shoulder does not seem to favor any single coordinate frame for position sense. We also looked at the potential accuracy in detecting changes in joint angles based on the distribution of muscle spindles throughout the human body. The distribution of spindles about individual joints accounts well for psychophysical data showing a proximodistal descending gradient of angular resolution that partially reflects the relative importance of more proximal joints for determining the location of the end-point.

An important aspect of motor function is our ability to rapidly generate goal-directed corrections for disturbances to the limb or behavioural goal. Primary motor cortex (M1) is a key region involved in feedback processing, yet we know little about how different sources of feedback are processed by M1. We examined feedback-related activity in M1 to compare how different sources (visual versus proprioceptive) and types of information (limb versus goal) are represented. We found sensory feedback had a broad influence on M1 activity with ~73% of neurons responding to at least one of the feedback sources. Information was also organized such that limb and goal feedback targeted the same neurons and evoked similar responses at the single-neuron and population levels indicating a strong convergence of feedback sources in M1.

200 27 Main Text: 6285 28 Figure Captions: 844 29 30 31 Abstract 47 48Modern control theory highlights strategies that consider a range of factors, such 49 as errors caused by environmental disturbances or inaccurate estimates of body or 50 environmental dynamics. Here we reveal similar diversity in how humans naturally 51 adapt and control their arm movements. We divided participants into groups based 52 on how well they adapted to interaction loads during a single session of reaching 53 movements. This classification revealed differences in how participants controlled 54 their movements and responded to mechanical perturbations. Interestingly, 55 variation in behaviour across good and partial adapters resembled simulations from 56 stochastic and robust optimal feedback control, respectively, where the latter 57 minimizes the effect of disturbances, including those introduced by inaccurate 58 internal models of movement dynamics. In a second experiment, we varied the 59 interaction loads over short time periods making it difficult to adapt. Under these 60 conditions, participants who otherwise adapted well altered their behaviour and 61 more closely resembled those using a robust control strategy. Taken together, the 62 results suggest the diversity of how humans control and adapt their arm movements 63 may reflect the accuracy of (or confidence in) their internal models. Our findings 64 may open novel perspectives for interpreting motor behaviour in uncertain 65 environments, or when neurologic dysfunction compromises motor adaptation. 66 67 68 69 70 how the controllers handle errors and uncertainty in their internal models of the 107 body and environment. Taken together, the results and simulations illustrate how 108 the accuracy of (or confidence in) internal models of the body and environment may 109 influence the control strategies individuals select to interact with their environment. 110 111 112 113 114 115 116 5 Results 117 Experiment 1. Learning to move in the presence of novel interaction loads 118We examined how healthy participants (N = 40; right handed) adapted their 119 reaching movements when exposed to novel interaction loads. Participants first 120 completed unloaded movements (Baseline) to three targets presented in random 121 order (T1-T3, radius = 1 cm; Fig.1A). We positioned the targets so that the arm 122 displacements required to reach them were identical across participants. Target 1 123 required shoulder motion (T1), Target 2 required combined shoulder and elbow 124 motion (T2), and Target 3 required elbow motion (T3). We then introduced novel 125 shoulder-joint loads ( Fig.1 A-B; Adaptation) that were proportional to the angular 126 velocity of the elbow joint (i.e., interaction loads) 24 . The interaction loads 127 systematically disturbed the participant's arm while they reached the training 128 targets (T2-T3). Participants were instructed to reach the targets within 450-650 129 ms. Movement durations were determined as the time between when the 130 participant's hand left the start ...