Answers are grounded in this paper and cite their sources.
Nonnegative signal and background components, bounded observational error, constraints on background smoothness and band-free anchor measurements define a set of admissible decompositions. Linear-fractional optimization gives the exact minimum and maximum band ratios over this set, while a feasibility problem tests consistency of the imposed constraints. A closed-form bound relates remaining area uncertainty to the anchor spacing, background curvature, noise and band profile. In a study of 320 simulated spectra, the computed intervals contained the known ratios whenever the imposed assumptions were satisfied. Across 3,282 additional simulated configurations, the analytical bound was respected; its finite values were con-servative. Under bounded-noise simulations, restricted central integration windows improved identification for the investigated profiles, particularly for Lorentzian bands; on measured backgrounds, this advantage appeared only at high signal-to-noise ratio. In a semi-synthetic study using 70 measured sulfate-only spectra, leave-one-concentration-out calibration produced intervals containing the known ratios in all 2,520 evaluated window and sensitivity configurations. However, precise identification was often unavailable despite close agreement among six conventional baseline-correction settings. Application to measured nitrate–nitrite mixtures was further complicated by nitrate-band overlap, and the available calibration constraints did not identify the corrected ratios tightly. These results distinguish numerical agreement among baseline estimates from identification supported by explicitly defined spectral assumptions.
No takes yet. Share an insight, caveat, or question.
Seyedeh Rahemeh Amiri (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: