Kernel Flow (v1) is a TD-fNIRS (time-domain function near-infrared spectroscopy) wearable headset, consisting of 52 modules (52 sources, 312 detectors), with a sampling rate of 200Hz & 500+ channels. The modules are built around miniaturized laser drivers, custom integrated circuits, and specialized detectors.
The modules are organized into 4 plates on each side of the head (8 total), covering the frontal, parietal, temporal, and occipital lobes. Each module consists of a central dual wavelength laser source surrounded by 6 hexagonally arranged detectors. Light is transmitted from source or to detector locations using spring loaded light pipes.
I was the responsible engineer for developing the optical train and the entirety of the optical module mechanical sub-assemblies and assemblies. I developed the Zemax models for the source (imaging + illumination) and detector optics (illumination), as well as the thermal simulations to note the expected source & detector temperatures. Additionally, I also setup & performed non-sequential monte-carlo simulations to model the light performance through bulk scattering tissue based on various anisotropy & index values pulled from literature.
Working with the modeling team, I tessellated the modules on the head to cover certain regions of interest, and developed a wearable head-set to hold the modules. I developed the raw surface that was modeled after various NIOSH head-forms, and acted as the foundation surface for creating the module plates, as well as the outer decorative plates for the system.
Date
March 2019, 2021
Designer
Isai Olvera
