Research

Systems engineering of human-use adaptive Deep Brain Stimulation systems for Essential Tremor and Parkinson’s Disease: In my work with the Activa PC+S and Nexus-D, I have lead technical and algorithm development in early demonstrations of adaptive DBS within human patients with both Essential Tremor (ET) and Parkinson’s disease (PD). My work represents some of the earliest adaptive deep brain stimulation studies in humans with a chronic implant, and I have published a variety of systems and algorithms that were used in conjunction with the Activa PC+S and Nexus-D to expand the understanding of what is possible from a technical perspective in the field. I have demonstrated the use of inertial sensors, electromyography, and implanted local-field potentials as signal sources for these closed-loop systems. I have also examined algorithmic design for adaptive deep brain stimulation in the form of adjusting stimulation to reduce ET and PD tremor based off both direct and indirect tremor biomarkers, using either the tremor intensity itself as a control parameter or anticipating tremor by identifying when a patient is likely at risk of experiencing symptoms. Using my knowledge in control theory I aided Helen Bronte-Stewart at Stanford through the identification a key need in order to translate her proposed adaptive DBS algorithm to practice, which was to include a ‘dead-band’ where no control action would be taken to allow for natural beta oscillations to occur within a physiologically ‘normal’ range. This algorithm was described in more detail within a patent application filed by Medtronic that I was included on given my development contribution, illustrating that the algorithms developed by my work may provide near-term benefit to patients through industrial translation. By taking a tools-based approach to the development of my Activa PC+S and Nexus-D experiments, my graduate student peers were able to adapt my work to support ongoing investigations into adaptive programming and gamma-band based control within Phil Starr’s lab at UCSF, illustrating the value of collaboration and producing generic research tool software packages in the growing area of chronic adaptive neuromodulation. Several notable publications based upon the Activa PC+S and Nexus-D work I have done include:

  1. Herron, M. Thompson, T. Brown, H.J. Chizeck, J. Ojemann, A. L. Ko, “Cortical brain computer interface for closed-loop deep brain stimulation”, IEEE Transactions on Neural Systems and Rehabilitation Engineering, 25.11 (2017): 2180-2187
  2. Herron, M. Thompson, T. Brown, H.J. Chizeck, J. Ojemann, A. L. Ko, “Chronic ECoG Sensing for Movement Intention and Closed-Loop DBS with Wearable Sensors in an Essential Tremor Patient”, Journal of Neurosurgery, 127.3 (2017): 580-587
  3. Malekmohammadi, J. Herron, A. Velisar, Z. Blumenfeld, M.Trager, H. J. Chizeck, H. Brontë‐Stewart, “Kinematic Adaptive Deep Brain Stimulation for Resting Tremor in Parkinson’s Disease”, Movement disorders: official journal of the Movement Disorder Society, 31.3 Jan, 2016, pp 426-428

Upon my return to the University of Washington as an Assistant Professor in Neurological Surgery, I was excited to be able to reestablish my involvement in these collaborations, but now as a mentor and advisor for the graduate students that followed me as they work on their own PhD projects. Based on my experience in industry at Medtronic, I was more keenly aware of the translational barriers that lay ahead for adaptive and closed-loop neuromodulation methods. Specifically, I am primarily concerned with the future burden these systems may pose to clinicians due to the sheer complexity of configuring closed-loop algorithms. While many of the clinicians I work with are capable of tuning embedded algorithms for their patients, it requires both well-developed domains of expert knowledge that most clinicians currently do not have as well as a significant time investment into each patient that many practices will not be able to afford. For this reason I have focused my own collaborative efforts and my mentored student’s time towards addressing the clinical burden these systems pose by a) developing automated algorithm optimization methods that remove the complexity burden from the clinician: my PhD student Benjamin Ferleger (co-advised by Prof. Howard Chizeck) evaluated automated classifier optimization methods for embedded Activa PC+S stimulation algorithms for our Essential Tremor patients; b) explore alternative algorithms that require less patient specific configuration on the part of the clinician: working with Helen Bronte-Stewart’s team at Stanford, I have been a remote technical mentor and contributor to ongoing work evaluating alternative gait-based kinematic algorithms to improve aDBS treatment of freezing of gait in PD; and c) investigate methods for automated evaluation of mobile-based symptom assessment data: my recent Master’s student Kazi (Sabrina) Sonnet published several papers illustrating methods for classifying spiral drawing task data in healthy, DBS-On, and DBS-Off cases. Notable published work in these efforts include:

  1. Ferleger, B. Houston, M. Thompson, S. Cooper, K. Sonnet, A. Ko, J. Herron, H. Chizeck, “Fully implanted adaptive deep brain stimulation in freely moving essential tremor patients” Journal of Neural Engineering 17.5 (2020): 056026
  2. O’Day, Y. Kehnemouyi, M. Petrucci, R. Anderson, J. Herron, H. Bronte-Stewart, “Demonstration of Kinematic-Based Closed-Loop Deep Brain Stimulation for Mitigating Freezing of Gait in People with Parkinson’s Disease” 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Montreal, Canada, July 2020
  3. Sonnet, B. Ferleger, A. Ko, H. Chizeck, J. Herron, “Multi-class classification and feature analysis for FTM drawing tasks in a digital assessment of tremor”, 20th IEEE International Conference on BioInformatics and BioEngineering (BIBE), Virtual Conference, Oct 2020

Medtronic Summit RC+S software development to support next-generation adaptive Deep Brain Stimulation research: After completing my Ph.D., I began working within the Medtronic developing the next generation of neuromodulation research tools. The next generation of human-use research tools being developed was the Olympus RC+S implantable device as part of the Summit System, which also includes a package of software and hardware aimed to push the boundaries of chronically-implanted neuromodulation research far beyond what Activa PC+S and Nexus-D was capable of. Based on my experience as a heavy Activa PC+S and Nexus-D experimentalist, I led the development effort around the system’s application programming interface (API) software that researchers would use to write their own protocol-specific applications. The APIs developed enable neuromodulation researchers to develop a research application that can both perform 1) computer-in-the-loop stimulation for in-clinic testing of new sensors and algorithms, and 2) allows for the configuration of embedded algorithms in the device so clinician researchers can evaluate how potential algorithms function in chronic usage. This API also handles the logging of all application interactions with the INS in a known logging format in such a way to enable cloud-based data storage and analysis to support new science discovery by examining Summit data across patients and devices that was not possible in the past. While the Summit System has not yet been implanted in humans (though the master file has been submitted and an investigational-device exception (IDE) has been approved by the FDA), it has been widely implanted in animal models for preclinical testing and early scientific discovery. Through this preclinical work, I have collaborated extensively with Mayo clinic researchers Greg Worrell and Ben Brinkmann who have used the Summit System to support their epilepsy research in dogs (and who later plan to implant humans now that their IDE is approved). While the Summit System is still quite new, the team has already had the opportunity to publish about the technological innovations and scientific discoveries enabled with the Summit System, several examples with are shown below:

  1. Herron, S. Stanslaski, T. Chouinard, R. Corey, H. Orser, T. Denison, “Bi-directional Brain Interfacing Instrumentation”, 2018 IEEE International Instrumentation and Measurement Conference (I2MTC), Houston, USA, May 2018
  2. Stanslaski, J. Herron, E. Ferhmann, R. Corey, H. Orser, T. Adamski, T. Denison, “Creating Neural ‘Co-Processors’ to Explore Treatment for Neurological Disorders”, 2018 IEEE International Solid-State Circuits Conference (ISSCC), San Francisco, USA, Feb. 2018
  3. Kremen, B. Brinkmann, I. Kim, S. Chang, J. Herron, S. Baldassano, E. Patterson, B. Litt, T. Denison, G. Worrell, “Continuous active probing and modulation of neural networks with a wireless implantable system”, 13th IEEE Biomedical Circuits and Systems Conference (BIOCAS), Turin, Italy, Oct. 2017

After leaving Medtronic to rejoin the team at the University of Washington in 2019, I was keenly interested in staying heavily involved in research with the Summit RC+S system. My unmatched knowledge of the workings of the Summit API provides me with deep insight into how to configure and work with the system for maximal effectiveness in a research study and also being able to easily identify sources of error within the sophisticated Summit System. This Summit System expertise allowed me to easily establish collaborations with Summit-focused investigational sites: a) I am engaged in technological development with Helen Bronte-Stewart in her study to evaluate new aDBS methods for PD based on beta-bursting activity within the STN; b) establishing a collaboration with Drs. Phil Starr and Simon Little at UCSF which expands at-home data collection in their Summit study through a newly funded project by the Weill Neurohub; c) working closely with the NIH-funded Open Mind Consortium to develop new device and coding-language agnostic software development tools, in particularly close collaboration with David Borton of Brown University. Initial publications in these early collaborations include:

  1. Petrucci, R. Anderson, J. O’Day, Y. Kehnemouyi, J. Herron, H. Bronte-Stewart, “A Closed-Loop Deep Brain Stimulation Approach for Mitigating Burst Durations in People with Parkinson’s Disease”, 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Montreal, Canada, July 2020
  2. Roarr, R. Perrone, F. Jamshed, R. Gilron, T. Denison, P. Starr, J. Herron, D. Borton, “OMNI: Open Mind Neuromodulation Interface for accelerated research and discovery” accepted to 10th International IEEE EMBS Conference on Neural Engineering (NER’21), Virtual Conference, May 2021