Bayesian maximum likelihood estimator of phase retardation for quantitative polarization-sensitive optical coherence...

Area of Science:

• Biomedical Optics
• Ophthalmology
• Medical Imaging

Background:

• Conventional birefringence estimation in polarization sensitive optical coherence tomography (PS-OCT) is prone to significant bias due to system noise.
• Accurate measurement of local phase retardation (birefringence) is crucial for understanding tissue optical properties.

Purpose of the Study:

• To develop and implement a maximum likelihood estimator (MLE) for accurate birefringence measurement using Jones-matrix-based PS-OCT.
• To overcome the limitations of conventional mean estimation methods in the presence of noise.

Main Methods:

• The estimator was designed using Bayes' rule to model the relationship between measured and true birefringence distributions.
• A Monte-Carlo method was employed to compute the likelihood function for numerical implementation.
• The proposed estimator was validated through numerical simulations and experimental data.

Main Results:

• The MLE demonstrated asymptotic unbiasedness, even under low signal-to-noise ratio conditions.
• The estimator performed accurately for phase retardations near the measurement range edges.
• Application to in vivo human anterior eye imaging revealed detailed clinical features.

Conclusions:

• The developed maximum likelihood estimator significantly improves the accuracy of birefringence measurements in PS-OCT.
• This technique offers enhanced diagnostic capabilities for ocular tissues by providing detailed structural and optical information.
• The MLE is robust against noise and suitable for clinical applications in ophthalmology.