Research

research directions

We seek to advance the fields of skeletal muscle neural control and biomechanics, ultimately to improve human-robot physical interaction. We combine multi-scale neuro-mechanical skeletal muscle modelling with real-time human-robot interfacing for creating novel model-based control technologies for skeletal muscle training, movement enhancement and rehabilitation. 

Multi-scale skeletal muscle neuromechanics in vivo We establish clinically viable interfaces with the human central nervous system that give access to the function of neural cells, such as spinal motor neurons. We build subject-specific multi-scale models of the human neuro-musculo-skeletal system, with an emphasis on skeletal muscles and series elastic tendons, that can translate neural recordings into accurate predictions of the resulting mechanical function in vivo in the intact moving human.

Real-time neuromechanical modelling Current clinical biomechanics involves lengthy data acquisition and time-consuming offline analyses. We develop innovative methods for the accurate analysis of skeletal muscle neuromechanics in vivo in the intact human in real-time. This will enable development of radically new healthcare technologies for the continuous monitoring of neuro-muscular function and remodelling, real-time bio-feedback of internal mechanical forces as well as neuro-muscular interfaced robots for motor training.

Exoskeleton model-based control We develop advanced online musculoskeletal modeling schemes that predict how an individual’s neuromusculoskeletal system responds to wearable devices connected in parallels to their residual limbs over time. We use dynamic simulation to predict how an individual's neuro-muscular system responds and adapt during long-term interaction with complex robotic exoskeletons. This information is used in real-time to create new paradigms of model-based control for wearable robots that can restore or enhance human motor capacity via personalized closed-loop predictive training.

Prosthesis model-based myoelectric control We define and experimentally test novel man-machine interfaces based on the discharge timings of spinal motor neurons and accurate predictions of the emerging physical behavior at the musculoskeletal level. This leads to novel mode-based myoelectric control schemes for bionic limbs. The development of man-machine interfaces that account for an individual’s neuromusculoskeletal system will open unprecedented opportunities to address clinically relevant rehabilitation challenges via biomimetic wearable assistive technologies.

Motor augmentation technologies We develop and test various techniques for the intention detection of people with Duchenne Muscular Dystrophy (DMD) and their integration with active upper extremity assistive devices. We aim to translate these technologies to the users and that is why we aim for user centered developments both for the design of devices and the intention detection.

3D tissue engineered skeletal muscles in vitro We develop platform for culturing 3D skeletal muscles in vitro and perform long-term trainings (> 20 days). By analysing bioreactor's cantilever/pillar deformation, tissue's filament alignment and proteins expression, we study tissue structural remodelling as well as change in tissue contractile capacity. We use this in vitro technique to create accurate in silico models of skeletal muscle remodelling that can be merged within numerical formulations that can predict skeletal muscle form and function in vivo.

Our Software We develop open source software to gain deeper insights on how the human nervous system interacts with the musculoskeletal system at mjltiple spatio-temporal scales. We apply this software also for the development of novel human-robot interfaces.