- #1

- 1

- 0

## Main Question or Discussion Point

Hello,

I don't know if anyone is familiar with OCT, but my question is rather specific and a well explained answer is in my opinion nowhere to be found ...

In OCT, a low coherent light source is used in a simple Michelson interferometer setup. One beam (reference beam) is sent onto a known path length with a mirror on the end, the other beam (sample beam) is reflected by scatterers inside the sample. The two beams come back and they form an interference pattern only if the sample beam has travelled a distance that can only differ from the known reference length no more than the coherence length of the source. By scanning the reference path by moving the mirror, you can scan the whole depth of the sample and get an A-scan.

This setup is called Time Domain OCT.

My question concerns another kind of setup, which is newer: Fourier Domain OCT.

In FD OCT, the reference mirror and thus the reference path length is keeped fixed. It relies on the Wiener-Kintchin relation between the spectral density of the signal and the autocorrelation function (the interference). It basicly measures the present frequency modulations in the spectrum and from this an inverse Fourier Transformation delivers depth information for some magical reason. I don't know why.

I also can't understand how interference can occur between the sample beam and the reference beam over the whole length of the sample (which is millimeters and falls way out of the coherence length of microns). I thought the whole point was just that interference can only occur within the coherence length and therefore interference between two beams that have travelled a different (outside the coherence length) distance can never interfere...

I know this is quite specific, but maybe there is an OCT guru out there, who happens to read this question

Thanks

I don't know if anyone is familiar with OCT, but my question is rather specific and a well explained answer is in my opinion nowhere to be found ...

In OCT, a low coherent light source is used in a simple Michelson interferometer setup. One beam (reference beam) is sent onto a known path length with a mirror on the end, the other beam (sample beam) is reflected by scatterers inside the sample. The two beams come back and they form an interference pattern only if the sample beam has travelled a distance that can only differ from the known reference length no more than the coherence length of the source. By scanning the reference path by moving the mirror, you can scan the whole depth of the sample and get an A-scan.

This setup is called Time Domain OCT.

My question concerns another kind of setup, which is newer: Fourier Domain OCT.

In FD OCT, the reference mirror and thus the reference path length is keeped fixed. It relies on the Wiener-Kintchin relation between the spectral density of the signal and the autocorrelation function (the interference). It basicly measures the present frequency modulations in the spectrum and from this an inverse Fourier Transformation delivers depth information for some magical reason. I don't know why.

I also can't understand how interference can occur between the sample beam and the reference beam over the whole length of the sample (which is millimeters and falls way out of the coherence length of microns). I thought the whole point was just that interference can only occur within the coherence length and therefore interference between two beams that have travelled a different (outside the coherence length) distance can never interfere...

I know this is quite specific, but maybe there is an OCT guru out there, who happens to read this question

Thanks