In-vivo classification of THC and CBD contents in cannabis plants using hand-handled Raman spectrometer

We present an innovative approach to the in-vivo classification of THC/CBD-rich
cannabis plants through a novel miniaturization strategy for Raman spectroscopy. By
developing compact Raman spectrometers that utilize patented technology based on cheap
non-stabilized laser diodes, densely-packed optics, and small pixel size sensors without
cooling, the study achieves performance comparable to more expensive, research-grade
systems. This miniaturization is facilitated by real-time calibration of Raman shift and
intensity using a built-in reference channel. The miniRaman spectrometer effectively
records high-quality Raman spectra of fresh cannabis and its products without the need for
sample or environment preparation, identifying characteristic peaks of primary
phytocannabinoids such as THC, CBD, and CGB and avoiding time-consuming HPLC
analysis.
Through spectral deconvolution and chemometrics, quantitative analysis becomes
possible, significantly reducing the influence of fluorescence for more precise analysis [1].
The application of this technology allows for the identification of THC or CBD-rich plants
with a high accuracy rate of 92%, demonstrating the potential of Raman spectroscopy
aided by machine learning for rapid, non-destructive cannabis classification.