NOVEL APPROACHES TO QUANTIFYING, TRACKING AND ENHANCING THE PERFORMANCE OF THE ELECTRODE-TISSUE INTERFACE IN MICROWIRE BRAIN IMPLANTS

Open Access
- Author:
- Paralikar, Kunal J
- Graduate Program:
- Bioengineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- November 24, 2008
- Committee Members:
- Ryan Clement, Dissertation Advisor/Co-Advisor
Ryan Clement, Committee Chair/Co-Chair
Andrew Webb, Committee Member
Alistair J Barber, Committee Member
Nadine Barrie Smith, Committee Member - Keywords:
- microelectrode array
microelectrode insertion force
collagenase
neural implants
brain-machine interface
common average referencing
inter-electrode correlation
common-noise
multi-unit recording
Spike detection
Biomedical MRI
neurophysiology
MRI heating
Immunohistology - Abstract:
- Intracortical microwire electrodes record extracellular multi-unit neural activity that is essential to drive neuroprosthesis systems and to develop our understanding of brain functions like learning and memory. Implantation of electrode arrays in the brain is a complex and invasive undertaking that requires excision of skull bone, removal of dura mater and piercing of pia mater meningeal layer. Typically, neuronal spiking events are detected by their relatively large amplitudes as compared to the rest of the signal or by their characteristic shapes. In order to improve microelectrode’s recording performance, researchers have experimented with multiple electrode substrates, insulation materials and implant strategies. However, all the currently available electrode systems have limited clinical applicability because they produce highly variable recording outcomes and most fail to record neural activity beyond one year. It has been hypothesized that this variability is modulated by implantation trauma and tissue response associated with presence of foreign material in the neural tissue. Development of optimal electrode structures and implant strategies has been impeded by an absence of objective techniques to quantify and monitor the state of tissue-electrode interface in order to understand its degradation and effectively compare different electrode platforms. This dissertation documents the development of novel approaches to help the processes of implantation, neuronal spike detection-quantification, and real-time interface monitoring. It reports an objective and innovative algorithm that utilizes information on multiple electrodes of an array to improve neuronal spike detection and measurement that can be used to quantify the recording performance of the interface. In doing so it also achieves significant reduction is downstream data processing. Minor addition to standard implantation techniques is reported that uses collagenase enzyme enabled disruption of the meningeal layer in order to reduce implantation trauma and electrode insertion forces as well as improve chronic recording. The technique can also help in the development and deployment of flexible biocompatible electrodes. In the last part of this dissertation, the safety of employing magnetic resonance imaging as a non-invasive imaging modality for real-time tracking of the longitudinal tissue changes around the electrode is demonstrated. It is shown that neural firing events are consistently detected even after numerous imaging sessions thereby pointing to the viability of the tissue close to the uninsulated recording site of the microelectrode. Taken together, these contributions lay the ground-work for a multitude of studies to realistically measure electrode performance and better understand interface degradation thereby enabling improved identification and incorporation of modifications to electrode structure, implantation and monitoring processes with the aim of extending microelectrode functionality.