Abstract
Coded aperture spectroscopy allows for sources of large étendue to be efficiently coupled into dispersive spectrometers by replacing the traditional input slit with a patterned mask. We describe a coded aperture spectrometer optimized for Raman spectroscopy of diffuse sources, (e.g., tissue). We provide design details of the Raman system, along with quantitative estimation results for ethanol at non-toxic levels in a lipid tissue phantom. With 60 mW of excitation power at 808 nm, leave-one-out and blind cross-validation of partial least squares (PLS) regression models achieve r2 > 0.98. Leave-one-out cross-validation demonstrates prediction errors of < 15% at the common legal limit for intoxication (17.4 mmol/L = 0.08% by vol) and the best blind cross-validation achieves < 12% error at this concentration.
Get full access to this article
View all access options for this article.
