A developmental electro-optical system would enable remote real-time viewing of a scene through a multimode optical fiber, as though the fiber were a conventional image-transmitting optic like a lens or prism. The system is intended to be a prototype of fiber-optic imaging probes that could be made very thin and could be particularly useful as minimally invasive probes in medical diagnosis.

A previous version of the system was described in "Multimode Optical Fiber as Imaging Probe" (NPO-19671), NASA Tech Briefs, Vol. 22, No. 2 (February 1998), page 21a. Both versions were conceived to address the following issue: It would be desirable to use an optical fiber as an imaging probe or imaging optic. Ordinarily, it would not be possible to do this because image information would become distorted (scrambled) during propagation along the fiber. The developmental system would make it possible by exploiting phase conjugation to reverse the scrambling.

A Hologram Would Be Generated in the Crystal at C during calibration. During readout, the object at A would be viewed at B by exploiting phase conjugation at C to reverse the scrambling that occurs during propagation along the optical fibers. The microelectromechanical Fabry-Perot interferometer would make it unnecessary to replace the object with a mirror during calibration.

The system (see figure) would include two multimode optical fibers, which would be terminated side by side at one end (location A) facing an object that one seeks to view. In the previous version, the tips of the fibers would face the object directly across a gap of width s. In the present version, a microelectromechanical Fabry-Perot interferometer (for use as described below) would be interposed between the tips of the fibers and the gap s. To ensure the coupling of sufficient phase information into and out of the fibers, the gap must very narrow; specifically, s << D2/ λ, where D is the aperture diameter of a fiber and λ is the wavelength of the light.

A source of light (in this case, a laser) at location C would illuminate the object via fiber 2. An observer at location B would attempt to view the illuminated object through fiber 1. The problem is to predistort the illumination (prescramble the amplitudes and phases of the fiber-optic waveguide modes of the illuminating electromagnetic field) in such a way as to compensate for the scrambling that occurs during transmission of the image along fiber 1 from A to B, so that the image of the object would arrive unscrambled at B.

The solution would involve the generation and use of a hologram in a phase-conjugating crystal (in this case, a photorefractive crystal) at location C during a calibration phase of operation. First, a flat mirror would be placed facing the tips of the optical fibers at location A, where the object would later be placed for viewing. A source of light would be placed at B, where the observer would later be stationed. Light from this source would travel through fiber 1 to A, where it would be reflected into fiber 2. Upon emerging from fiber 2 at C, the light would enter the crystal. At the same time, the crystal would be illuminated with a reference (plane-wave) beam of light. Interference between the reference beam and the light emerging from fiber 2 would produce the desired hologram via the photorefractive effect. The hologram would encode the information about scrambling in both fibers 1 and 2.

Once the hologram had been generated, one could exploit the phase-conjugation principle to reverse the propagation of the optical signal and thus reverse scrambling during a readout phase of operation. The crystal would be illuminated with the phase conjugate of the reference beam (in essence, a beam of the same wavelength propagating along precisely the reverse of the path of the reference beam); this would cause reverse-propagation with unscrambling of light from C back to A, then back to B. If the mirror were replaced by the object to be viewed, then the reverse-propagating light would illuminate the object and the image of the object would spatially modulate the reverse-propagating beam, such that an undistorted image of the object would appear, upon completion of reverse propagation and unscrambling, at B.

One obstacle to implementation is that exposure of the crystal to the reverse reference beam during the readout phase of operation would gradually erase the hologram. Thus, the photorefractive crystal material should be one that has a long characteristic erasure time, and one must complete the readout well within that time. The prime candidate photorefractive material is BaTiO3, in which erasure lags by 1 to 20 seconds.

Another obstacle to implementation is the need to place the flat mirror and then the object of interest at location A. In a real-time-viewing application, it would be necessary to switch repeatedly between the flat mirror and the object of interest in order to alternate the calibration phase with the readout phase. The microelectromechanical Fabry-Perot interferometer would make it possible to leave the object of interest in place and would make it unnecessary to use a separate mirror. Instead, the interferometer could be switched electrically between (1) a state of total reflection, in which the interferometer itself would serve as the mirror for calibration; and (2) a state of total transmission, in which light would be coupled between the fibers and the object of interest during readout. An additional advantage of this concept is that microelectromechanical Fabry-Perot interferometers could be fabricated at low cost.

This work was done by Deborah Jackson of Caltech for NASA's Jet Propulsion Laboratory.


This Brief includes a Technical Support Package (TSP).
Development of multimode-optical-fiber imaging probe

(reference NPO20429) is currently available for download from the TSP library.

Don't have an account? Sign up here.

Photonics Tech Briefs Magazine

This article first appeared in the March, 2000 issue of Photonics Tech Briefs Magazine.

Read more articles from the archives here.