E14-633 (3pm)
Rewiring Neural Conduits.
Engineering Neuromuscular Tissues for Bidirectional Neuroprosthetic Interfacing
Contemporary technological approaches to address limb loss and neuromuscular dysfunctionconsist of synthetic, mechanical devices which lack an intimate bidirectional interface with nervous tissues. On the therapeutic front, the current amputation paradigm disrupts neuromuscular architecture, discards sensory organs and provides no anatomical or prosthetic replacement. This precludes the generation of afferent sensory feedback, which is critical for sensory integration, motor planning, peripheral and central neurological health, and myoelectric prosthesis control. Utilizing a paradigm of coevolution, I simultaneously engineer neuromuscular anatomy and bioelectronics to enable seamless, bidirectional neuroprosthetic interfacing.
In this dissertation, I describe the design and preclinical validation of the regenerative agonist-antagonist myoneural interface (AMI) and myodermal interface (MI), which are reconstructive surgical models to restore musculotendinous and cutaneous sensory feedback, respectively. Then, through case-control studies, the functional outcomes of human subjects who have undergone below-knee and above-knee amputations incorporating native AMIs are compared to standard amputation controls. The effect of AMI amputation on sensorimotor neuroplasticity is investigated through anatomical and functional neuroimaging. These preclinical and clinical evaluations demonstrate the a) production of graded efferent and afferent signals, b) the maintenance of peripheral limb volume and central sensorimotor substrates, c) improvements in phantom sensation, phantom pain, and neuroprosthetic controllability, and d) decreased dependence on compensatory visuomotor circuitry. To address challenges with functional electrical stimulation (FES) of neuromusculature, employed for prosthetic feedback and control, I develop a closed-loop functional optogenetic stimulation system (FOS) for peripheral neuromuscular control. This system demonstrates greater accuracy, biomimetic orderly recruitment of fibers, and minimized fatigue during cyclic movements as compared to FES.
Spanning from animal models to human implementation, this dissertation presents 1) a model to design new surgical techniques for afferent/efferent signaling, 2) characterize the physiology following clinical translation, and 3) recursively apply the lessons to the design of neural interfaces back at the bench. In summary, the results of this work steer a shift of the clinical amputation paradigm towards one that performs strategic rewiring of neuromuscular constructs to enable improved neurological health and neural interfacing
Thesis Supervisor:
Hugh M. Herr, PhD
Professor of Media Arts and Sciences, MIT
Thesis Committee Chair:
Emery Brown, MD, PhD
Professor of Computational Neuroscience and Health Sciences and Technology, MIT
Thesis Reader:
Robert Langer, ScD
David H. Koch Institute Professor, MIT