Optical coherence tomography velocimetry based on decorrelation estimation of phasor pair ratios (DEPPAIR)

Maximilian G.O. Gräfe, Oleg Nadiarnykh, Johannes F. de Boer*

*Corresponding author for this work

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

Quantitative velocity estimations in optical coherence tomography requires the estimation of the axial and lateral flow components. Optical coherence tomography measures the depth resolved complex field reflected from a sample. While the axial velocity component can be determined from the Doppler shift or phase shift between a pair of consecutive measurements at the same location, the estimation of the lateral component for in vivo applications is still challenging. One approach to determine lateral velocity is multiple simultaneous measurements at different angles. In another approach the lateral component can be retrieved through repeated measurements at (nearly) the same location by an analysis of the decorrelation over time. In this paper we follow the latter approach. We describe a model for the complex field changes between consecutive measurements and use it to predict the uncertainties for amplitude-based, phase-based and complex algorithms. The uncertainty of the flow estimations follows from a statistical analysis and is determined by the number of available measurements and the applied analysis method. The model is verified in phantom measurements and the dynamic range of velocity estimations is investigated. We demonstrate that phase-based and complex (phasor) based lateral flow estimation methods are superior to amplitude-based algorithms.

Original languageEnglish
Pages (from-to)5470-5485
Number of pages16
JournalBiomedical Optics Express
Volume10
Issue number11
Early online date2 Oct 2019
DOIs
Publication statusPublished - 1 Nov 2019

Fingerprint

Dive into the research topics of 'Optical coherence tomography velocimetry based on decorrelation estimation of phasor pair ratios (DEPPAIR)'. Together they form a unique fingerprint.

Cite this