Translation and Engineering

Axis 1 of the lab's research, in depth. This page is for the engineer or translational scientist who wants to understand how the lab's mechanistic work informs device design, stimulation parameter selection, and clinical deployment.

The Failure-Mode Framework

Brain-machine interfaces fail in tissue, and the failures cluster into a recognizable set of mechanisms that materials science alone cannot explain or fix. The lab's framing of the field, developed across the "Why BCIs Fail" series and the underlying primary literature, identifies seven coupled failure cascades.

•         The first 30 seconds, mechanical insertion injury, vasculature disruption, and acute neuronal silencing

•         The stiffness mismatch, ongoing micromotion at the device-tissue interface and chronic mechanical injury

•         The healing response, which is the problem rather than the solution, foreign body response and glial encapsulation

•         The cells nobody was watching, oligodendrocytes, OPCs, mural cells, and the non-neuronal contributions to chronic device function

•         The energy crisis, metabolic and oxygen consumption limits on sustained neural firing under stimulation

•         The network problem, parametric stimulation drives circuit-level effects that are not predicted from single-cell models

•         We Were Wrong About What Electrical Stimulation Can Do, Network Inhibition

•         The Alzheimer’s-BCI Connection Nobody Talks About

•         The Trash Problem around implanted BCIs. Lysosomes are a cell's waste management system

•         Speed Wasn’t the Point. The mechanism is metabolic

•         Runway and ecosystem moats, the structural constraints on industry translation that materials and devices alone cannot overcome

The Forward-Translation Arm

The lab's mechanistic discoveries inform device design and clinical deployment through direct engagement with industry partners and clinical neuromodulation programs.

Examples of forward translation from the lab's work include the carbon-fiber microthread electrode design (Kozai et al., Nature Materials 2012), which informed subsequent ultrasmall electrode developments in the field, the pharmacological repositioning of clemastine for chronic recording stability (Chen, Cambi, Kozai, Biomaterials 2023), the demonstration that low-intensity pulsed ultrasound modulates microglial activation around chronic electrodes (Li et al., Nature Communications 2024), and the parametric framework for ICMS that resolves a 30-year question on perceptual fading by identifying neurovascular and metabolic constraints as the dominant mechanism.

The lab maintains active engagement with the neural interface industry through collaborations, advisory roles, and direct conversations with research scientists at major neural interface and medical device companies. The Fontis Biotechnology founding history (Dr. Kozai's graduate-school company) and the Neuralink co-founder offer (October 2016, declined on scientific grounds) anchor the lab's translational credibility.

Skills Developed by Trainees on the Engineering Track

•         Cellular and biophysical mechanism of device-tissue interactions

•         Parametric stimulation design and the mechanistic basis of stimulation parameter selection

•         Device characterization through in vivo electrophysiology, two-photon imaging, and computational modeling

•         Industry-relevant analytical pipelines, longitudinal recording analysis, unit tracking, statistical inference at appropriate sample sizes

•         Direct exposure to the industry research scientist role through internships, introductions, and post-graduation placement support