Dissecting muscle synergies in the task space

The research presented here concerns the notion of modularity and its role in the functional organisation of the human motor system. In Chapter 1, I introduce the concept of modularity and a popular computational approach to its investigation in the motor neurosciences known as muscle synergy analysis. I highlight open problems in this field, in particular the lack of a direct mapping of muscle synergies to task performance, and present the ways in which i will address these open problems both conceptually and analytically. In Chapter 2, I address current analytical limitations in the field by leveraging information- and network-theoretic tools to present a novel, generalisable approach to muscle synergy extraction under relaxed model assumptions. This approach builds on top of traditional methods and is referred to as the Network-Information Framework (NIF). In Chapter 3, I then employ the NIF to provide a new perspective on muscle synergies that is made implicit in this novel computational approach. This novel perspective integrates important findings from recent influential works showing how muscles not
only ’work together’ towards common task-goals as previously conceived, but also complementary and task-irrelevant objectives concomitantly. By directly including task parameters into muscle synergy extraction, i effectively dissect the task-relevant information dynamics underlying coordinated movement, thus providing a principled way to access this complex functional architecture. In Chapter 4, I further develop the NIF to simultaneously quantify diverse types of muscle interactions across inter- and intra-muscular scales, including functionally similar (i.e. redundant), -complementary (i.e. synergistic) and -independent (i.e. unique) interactions. In doing so, I reveal novel insights into movement control in health and with pathology. I also align current muscle synergy analysis with the forefront in theoretical understanding on human movement modularity. To conclude this work, in Chapter 5 I summarise these contributions, their implications for neurobiological mechanisms, and the novel research opportunities they present for the motor control field.