Intelligent Spine Interface
Julia Gangemi | Collaborators: Rhode Island Hospital, Providence VA Healthcare
At present, activity-based therapies are the only medical practices that can be used to enhance recovery after spinal cord injury (SCI). However, the most affected patients, who fail to produce active movements voluntarily, experience minimal benefits from such therapies. Several studies have now shown that spinal cord stimulation delivered at the right time can enhance a physical therapy rehabilitation program significantly, leading to restoration of volitional walking (when the stimulator is on). We therefore propose a bi-directional tool for sensing and stimulating to bridge a gap in the spinal cord, ‘reconnecting’ patches of eloquent nervous system tissue. The Intelligent Spine Interface (ISI) will be capable of reading and writing simultaneously to, and from, the human spinal cord both above, and below, the site of SCI. The ISI will interpret neural information from above a spinal cord lesion and transfer that information, via state-of-the-art artificial neural network-based interpreters, to sites below the lesion and restore volitional control of the lower limb. We will focus our therapeutic demonstrations on the restoration of bi-directional sensing and control of the legs and voluntary locomotion. The proposed technology is agnostic to the level of the spinal lesion and has far greater number of sites (electrodes) to interact with the nervous system, thus making the therapeutic potential far greater than current technologies.
Related Publications:
Smart, Safe, Switch
Gabriela Porto Machado | Collaborators: Providence VA Healthcare
Advances in neurotechnology promise new therapies for veterans and others living with spinal cord injury, chronic pain, and other neurological conditions, yet it can take decades for a device to move from the benchtop to the bedside. Today's implants are built from many components (amplifiers, stimulators, wireless telemetry, and signal-processing algorithms), and developers must choose between flexible lab systems that require major redesign before clinical use and specialized custom boards that lock in early design choices and add cost. This project creates a modular hardware and software development platform, running on a real-time operating system, in which components from different manufacturers plug in as interchangeable modules and signals are rerouted in software rather than through new circuit boards. Open, language-agnostic configuration tools let researchers control the system from any programming language, while each module can keep its developer's intellectual property protected. We are using the platform to build a fully implantable, bidirectional spinal cord interface that records spinal signals and adjusts stimulation in closed loop, validating it in long-term preclinical studies before clinical translation, and we will share the platform with the broader neurotechnology community to speed the development of new therapies.
3D Multicellular Models of the Brain
Sophie Brown , Abby Drexler , Claire Petersen , Serena Agarwal | Collaborators: Huang Lab
As the most common form of neurodegenerative disease, Alzheimer Disease (AD) contributes to nearly 60–70% of all cases of dementia globally. Currently, the path towards successful therapeutic intervention is obstructed by a limited understanding of disease etiology and the myriad of proposed underlying mechanisms. Evidence indicating increased levels of inflammatory markers in AD patients, as well as immune-associated loci of several AD risk genes, suggests that neuroinflammation is a contributing factor to pathogenesis. As microglia are dynamic mediators of neuroinflammation, determining their specific roles in AD pathology is essential to deciphering neuroinflammatory influence on underlying AD mechanisms. Here in the Borton Lab we are working to develop a hybrid in vitro model to study human iPSC-derived microglia (iMGs) in a 3D xenoculture system. We hypothesize that hybrid 3D xenocultures can be utilized to study human microglial behaviors that may contribute to AD pathogenesis in a physiologically-relevant in vitro environment. Our overall objectives in this research program are to (1) develop novel in vitro tools for studying complex cellular biology in the brain and (2) characterize the roles of human microglia during neuroinflammatory disease states.
Related Publications:
Understanding Cortical Control of Locomotion
Shivam Chaudhary , Aaron Gregoire , Alex Dawson-Elli , Gabriela Porto Machado
The field of brain-machine interfaces (BMI) for restoring forelimb motor function has made considerable progress over the last few decades. However, the development of hind-limb counterparts have remained relatively nascent. Although there have been recent advancements in restoring basic locomotion in both animal models and in the clinic, the number of developments for neuroprosthetics allowing direct control over voluntary hind-limb movements is still sparse. In this project, we are developing a closed-loop BMI system allowing for direct end-point control of the foot. Using implanted multi-electrode arrays and machine learning techniques, neural signals recorded from motor cortex will be decoded and translated into movement of a robotic actuator in real time. The viability and functionality of our system will be validated in a pedal positioning task. Additionally, in order to integrate both voluntary hind-limb movements with autonomous locomotion into one general-purpose BMI, it is necessary to understand how the motor cortex encodes hind-limb movement during these two behaviors. We employ an obstacle avoidance paradigm to probe the neural correlates and network dynamics of leg-M1 during both voluntary (e.g. originating from cortical areas) and autonomous (e.g. originating from spinal circuits) action. The ultimate goal would be to develop general-purpose, high-functioning neuroprosthetics for the hind-limb allowing for a wide variety of motor actions and enabling patients to regain full lower limb function.
Related Publications:
Neural Dynamics of Pain Processing in the Spine-Brain Continuum
Claire Petersen | Collaborators: Cleveland Clinic
This project focuses on the neural dynamics governing pain processing. Our goal is to develop a mechanistic framework of pain relay in order to optimize chronic pain therapies, such as spinal cord stimulation (SCS). While SCS has been used to treat chronic pain patients, it is unclear how stimulation modulates neural circuits that are responsible for pain perception. Therefore, we are developing biophysically realistic neural models of spinal and cortical circuits implicated in pain processing to examine the effects of electrical stimulation on neural activity. In parallel efforts, we are recording acute and chronic electrophysiology in mice to study spinal and cortical circuits in vivo in response to sensory stimuli.
Related Publications:
Neuromodulation therapy with Individualized Network Targeting in Depression
Anusha Allawala , Yunshu Fan , Stephanie Vartany , Michelle Akerman | Collaborators: Baylor College of Medicine, UCLA, Boston Scientific
Major Depressive Disorder (MDD) is the most common mental disorder in the US and the leading source of disability. Despite multiple treatment options, up to one-third of patients are diagnosed as having Treatment-Resistant Depression (TRD). While early clinical trials of Deep Brain Stimulation (DBS) for depression have shown promise, results from prior studies have been mixed due to a large number of factors, including the lack of understanding surrounding the disruption of circuitry in depression, and the mechanisms by which DBS perturbs these circuits. Additionally, TRD encompasses a wide variety of symptoms that vary within and across patients. We are using multi-modal approaches to individualize DBS therapy for patients with depression using multi-site DBS and intracranial recordings. Our goal is to ultimately tailor treatment for each patient depending on their network activity and the nature and severity of behavioral impairment.
Related Publications:
Responsive neuromodulation in Neuropsychiatric Illness
Ayan Waite , Michaela Alarie , Yunshu Fan | Collaborators: Baylor College of Medicine, Draper, Medtronic, University of Pittsburgh
Obsessive Compulsive Disorder (OCD) is a psychiatric illness marked by obsessions (recurrent unwanted or distressing thoughts) and compulsions (repetitive, ritualistic behaviors). OCD affects ~2% of the US population, and 10-20% of cases are treatment resistant. Deep Brain Stimulation (DBS) in the ventral capsule/ventral striatum (VC/VS) has been found to improve symptoms in approximately 50-70% of patients. While early trials of DBS have been promising, clinical trials have failed to date. These failures may be attributed to the “open-loop” nature of DBS, where stimulation parameters are chosen during infrequent visits to the clinician’s office. Further, the continuous stimulation fails to address the dynamic nature of OCD; symptoms often fluctuate over minutes to days. Titrating DBS to respond to symptoms as they arise (i.e. “Adaptive DBS”) may be a more effective approach for treating symptoms of OCD and reducing undesirable side effects of stimulation. We hope to design a closed-loop, adaptive system in which (1) electrodes would continuously record electrical activity from the brain, (2) recorded data would be used to classify maladaptive mental states as they arise, (3) and stimulation parameters would be adjusted accordingly to relieve symptoms.
Related Publications: