Answers are grounded in this paper and cite their sources.
This study presents a proof-of-concept monochrome camera system featuring a UV-sensitive sensor, enabling direct UV detection for non-destructive quantification of active pharmaceutical ingredient content in two-component canrenone-microcrystalline cellulose tablets. Differences in UV absorption between API and excipient resulted in distinct grayscale intensity distributions on the tablet surface, enabling active pharmaceutical ingredient (API) quantification. Tablets were prepared with API contents ranging from 0 to 6 w/w% and imaged in-line on a conveyor belt under 365 nm UV illumination. Acquired images were pre-processed through background elimination, and illumination correction, then grayscale intensity histograms were extracted. These histograms served as inputs for partial least squares regression, producing a linear correlation between predicted and nominal API content, with a root mean square error of calibration 0.20 w/w%, root mean square error of cross-validation 0.24 w/w%, root mean square error of prediction 0.27 w/w%. The obtained relative prediction error was 6.78% at a target of 4 w/w%. The approach demonstrates rapid, non-destructive assessment of API content without sample preparation. By combining a cost-effective direct UV-imaging method with multivariate analysis, the system shows potential for real-time monitoring of API content in continuous tablet manufacturing.
No takes yet. Share an insight, caveat, or question.
Molnar et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: