Thickness estimation with optical coherence tomography and statistical decision theory

Thickness estimation, which has a broad range of applications, plays an important role in the field of optical metrology. In this study, we investigate a new approach—combining optical coherence tomography (OCT) and statistical decision theory—for thickness estimation. We first discussed and quantified the intensity noise of three commonly used broadband sources, a super-continuum source, a super-luminescent diode (SLD), and a swept source. Furthermore, a maximum-likelihood (ML) estimator was implemented to interpret the OCT raw data. Based on the mathematical model and the ML estimator, simulations were set up to investigate the impact of different broadband sources in OCT for a thickness estimation task. We then validated the theoretical framework with physical phantoms. Results demonstrate unbiased nanometer-class thickness estimates with the ML estimator. The framework can be potentially used for film and surface shape metrology.