1. Field of the Invention
The invention relates generally to optical coherence tomography (OCT) and in particular to Fourier domain optical coherence tomography.
2. Description of Related Art
Optical Coherence Tomography (OCT) is a technology for performing high-resolution cross sectional imaging that can provide images of tissue structure on the micron scale in situ and in real time. In recent years, it has been demonstrated that Fourier domain OCT (FD-OCT), which so far employs either a wavelength swept source and a single detector or a broadband source and an array spectrometer, has significant advantages in both speed and signal-to-noise ratio as compared to time domain OCT (TD-OCT). In TD-OCT, the optical path length between the sample and reference arms needs to be mechanically scanned. For example, patent application WO03062802 (EP1470410, CA2474331, US20050018201) is a hybrid time and spectral domain OCT system in which a broadband source is combined with sub-depth range mechanical optical path length scanning and parallel detection of a set of optical spectral bands. This design can increase the signal-to-noise ratio and at the same time reduce the mechanical scanning range. But mechanical scanning is still required and the A-scan speed is thus limited.
In both swept source OCT (SS-OCT) and spectrometer-based spectral domain OCT (SD-OCT), the optical path length difference between the sample and reference arm is not mechanically scanned, instead, a full axial scan (also called A-scan) is obtained in parallel for all points along the sample axial line within a short time determined by the wavelength sweep rate of the swept source (in SS-OCT) or the line scan rate of the line scan camera (in SD-OCT). As a result, the speed for each axial scan can be substantially increased as compared to the mechanical scanning speed of TD-OCT and this is especially beneficial for real-time imaging of movable biological samples such as the human eye. In addition, SD-OCT and SS-OCT can provide substantially greater signal-to-noise ratio relative to TD-OCT, as explained by Mitsui and others (“Dynamic Range of Optical Reflectometry with Spectral Interferometry.” Japanese Journal of Applied Physics 38(10): 6133-6137). There are a number of patents as well as articles that either disclosed the basic concept of or discussed the advantages of Fourier domain OCT using either a swept single wavelength source combined with a single photodetector or a broadband source combined with an array spectrometer. Several of these articles and patents are listed separately under the REFERENCES section. These and other articles and patents cited are all incorporated herein as references of this invention.
However, these prior arts are based on purely employing either a swept single wavelength source combined with a single photodetector (thereafter called pure swept-source OCT or pure SS-OCT) or a broadband source combined with an array spectrometer, comprising an optical spectral dispersing element and an array of photodetectors (thereafter called pure spectral-domain OCT or pure SD-OCT). A pure SS-OCT or a pure SD-OCT system each has its advantages and disadvantages in terms of cost, speed, size, stability and other factors as will be elaborated shortly. Based on the advantages of FD-OCT and most importantly, the cost of currently available optical components, we describe an alternative FD-OCT system that not only retains the advantageous features of a pure SS-OCT and a pure SD-OCT, but also saves the cost of the overall system and increases the speed.
In order to fully appreciate the novel features of the present invention, let us first take a brief look at a pure SS-OCT system and a pure SD-OCT system.
A need therefore exists for an alternative FD-OCT design that can take the speed advantage of FD-OCT without requiring a high-speed array detector, or a high-speed wide-range swept laser source.
The sweep range and sweep speed of a continuously tunable laser are generally limited by the swept filter acting as a tuning element. In addition, laser dynamics limits the sweep speed, especially for longer-cavity lasers. (See for example, Huber, R. et al. (2005). “Amplified, frequency swept lasers for frequency domain reflectometry and OCT imaging: design and scaling principles.” Optics Express 13(9): 3513-3528).
Some examples of sweepable filters that can be used in tunable sources are: 1) piezoelectrically-driven Fabry-Perot filters, 2) unbalanced fiber-based Mach-Zehnder interferometers, 3) distributed Bragg reflectors (DBR), 4) distributed feed-back (DFB) in the gain medium of the laser itself, and 5) rotating gratings outside the laser cavity. For most filter types, especially the first four, smaller sweep ranges are more easily achieved than large sweep ranges. For example, a semiconductor-based DBR, tuned by carrier density, can change the refractive index by 1%, resulting in a tuning range of 1%. The desired tuning range for pure SS-OCT of biological samples is at least 25 nm, which is 2% of a typical laser wavelength of 1310 nm. Each of these filter types can be naturally adapted to pass several wavelengths simultaneously, because they are based on interference and can operate on multiple orders of interference simultaneously.
In this invention, a continuously swept multi-wavelength laser is combined with an optical multi-channel receiver. The multi-wavelength laser emits several optical frequencies or wavelengths simultaneously. This source covers a broad frequency range in a short time by sweeping the set of individual lasing frequencies so that each lasing frequency covers a portion of the full spectrum. The individual lasing optical frequencies are swept over a relatively small range between neighboring lasing frequencies. As a result, the cost of the FD-OCT system can be markedly reduced. The proposed continuously swept multi-wavelength laser can also be made at a low cost as it is only slightly different from standard tunable semiconductor lasers for telecom applications. An additional very beneficial feature of the invention is that the time required for each A-scan can now be substantially reduced, which means that the A-scan rate can be substantially increased relative to the single lasing wavelength swept-source. The individual lasing frequencies simultaneously excite the OCT interferometer, and a multi-channel optical receiver can separate and record the resulting interference signals from the individual optical frequencies. The required multi-channel receiver is now commercially available at a relatively low cost and the price is still continuously dropping as a result of the development for optical fiber telecom applications.
The present invention is an alternative Fourier domain optical coherence tomography (FD-OCT) system comprising a swept multi-wavelength laser, an optical interferometer and a multi-channel receiver (detector). The optical interferometer includes a beam splitter, which may be, for example, a fiber optic coupler, for separating the light along a sample and reference path. Light returning from both paths is combined and the combined light is measured by the detector.
The present invention also includes a method for performing Fourier domain optical coherence tomography, comprising the steps of simultaneously sweeping the wavelengths of a multi-wavelength laser, directing the light beam from the swept multi-wavelength laser to an interferometer, and detecting the interfered optical signal using a multi-channel receiver. By selecting a comb spacing of the multi-wavelength laser to be equal to or greater than the channel width of the multi-channel receiver and by sweeping each lasing wavelength across a spectral width equal to or greater than the laser comb spacing, each individual channel detector will record a partial spectral interferogram. The full spectral interferogram over a broad spectral bandwidth can be obtained by combining the data from the partial spectral interferograms.
A main object of the invention is to reduce the cost of an FD-OCT system and this is achieved by combining a swept multi-wavelength laser with a multi-channel receiver in an optical interferometer. Owing to the fact that the sweep range for each lasing wavelength of a multi-wavelength laser is substantially reduced as compared to that of a swept single wavelength laser that needs to cover the same overall spectral range, such a multi-wavelength laser can hence be fabricated at a relatively low cost as compared to the swept single wavelength laser. Meanwhile, one can also take advantage of the multi-channel receivers that have already been developed for fiber optical communications and are now commercially available at relatively low cost.
Another key advantage is the increase in the speed of each axial scan relative to pure SS-OCT. Since the tuning range for each wavelength of the multi-wavelength laser is substantially reduced, for a given a sweep rate limited by either the filter element or by laser dynamics, the time required to record one interferogram, corresponding to one axial scan, is also substantially reduced.
In some cases, the speed of the sweep over optical frequencies might be limited by the speed of the optical detectors or the digitizers recording the detected signals. In these cases, this invention has the advantage of using these detectors in parallel, with each detector required to record only a section of the full spectral interferogram within the time allowed to complete an axial scan.
A further advantage of this invention is that it can make more efficient use of the laser gain medium. Gain media designed to have broad gain bandwidth, such as quantum dot or dot-in-well based semiconductor optical gain media, are largely inhomogeneously broadened, meaning that distinct portions of the medium contribute to gain at different optical frequencies. All these distinct portions are often pumped simultaneously, such as in typical broadband semiconductor optical amplifiers (SOA) where a largely inhomogeneously broadened gain medium is excited using a single drive current (see for example, Yao J. et al. “Multiwavelength Erbium-Doped Fiber Ring Laser Incorporating an SOA-Based Phase Modulator” Photonics Technology Letters, IEEE 17(4): 756-758). In a frequency-swept laser then, some portions of the gain medium must maintain their population inversion for the majority of the sweep period during which these portions do not produce output light. The population inversion relaxes, typically less than one microsecond in a SOA, and must be continuously replenished, by current injection in an SOA. Allowing multiple optical frequencies to lase simultaneously increases the duty cycle during which each portion of the gain medium is producing output light, reducing the power requirements and heat dissipation of the source. Increasing the duty cycle also increases the ratio of lasing output to amplified stimulated emission, which is advantageous because amplified spontaneous emission in the laser output adds noise to the OCT image.
One object of the invention is to make use of existing lasers or adapt their fabrication technologies to the manufacturing of the tunable multi-wavelength laser. In this respect, the multi-wavelength laser can be made based on a sampled grating based semiconductor laser. Such a laser may have two sections, of which one is the lasing section incorporated with a sampled grating that can enable the sweeping of the multiple lasing wavelengths, and the other is an optical amplifier section that can compensate for the change in the output optical power. Alternatively, the tunable multi-wavelength laser can also be made from a tunable Fabry-Perot semiconductor cavity. Such a laser can also have two sections, of which one is a gain section for lasing and the other is a transparent section for sweeping the lasing wavelengths. A separate optical amplifier can also be made next to the laser to boost and also control the optical output power. Furthermore, the tunable multi-wavelength laser can also be made from a tunable ring or race-track semiconductor cavity. Such a laser can again have two sections, of which one is a gain section for lasing and the other is a transparent section for sweeping the lasing wavelengths. A separate optical amplifier can also be made next to the laser to boost and control the optical output powder.
Other types of laser source for the present invention include an extended multi-wavelength cavity laser that is made tunable with at least an optical amplifier and a tunable multi-wavelength filter functioning as part of the laser cavity. The extended long laser cavity is preferably made with optical fiber. Both the optical amplifier and tunable multi-wavelength filter are preferably either integrated optical waveguide based, or optical fiber based or optical fiber pigtailed.
Multi-channel receivers that can be used for the present invention include a monolithically integrated semiconductor multi-channel receiver, and a discrete optical demultiplexer or a series of thin film-DWDM Bandpass filters that can be combined with a series of photodetectors to function as the multi-channel receiver.
These and other features and advantages of the present invention will become more readily apparent to those skilled in the art upon review of the following detailed description of the preferred embodiments taken in conjunction with the accompanying drawings.
The processor combines the samples from the individual channels to form a full spectral interferogram and carries out a Fourier transform of the spectral interferogram to provide the information of the reflectance distribution along the depth within the sample 322, in a similar way as described by Huber, R. et al. (2005). “Amplified, frequency swept lasers for frequency domain reflectometry and OCT imaging: design and scaling principles” Optics Express 13(9): 3513-3528. Note that the reflectance detected includes not only specular reflections, but also other scattering processes that return light to the multi-channel receiver. In a preferred embodiment, the tunable multi-wavelength laser 302 simultaneously produces a comb spectrum 330 of n lasing wavelengths (λ1 to λn) that are all simultaneously swept across a range of optical frequency that is comparable to or greater than the channel spacing Δλ of the comb spectrum 330, and the channel spacing (also called the free spectral range) Δλ of the comb spectrum 330 is comparable to or greater than the individual channel width of the multi-channel receiver 310.
The tuning range now required for each lasing wavelength of the multi-wavelength laser 302 is reduced by approximately n times as compared to that of a tunable single wavelength laser that covers the same overall broadband spectral range. Each photodetector in the multi-channel receiver 310 will sequentially receive a partial spectral interferogram covering approximately a bandwidth of the corresponding channel. If the channels of the multi-channel receiver are spectrally spaced next to each other or more preferably, slightly overlapping, the overall spectral interferogram can be completely covered by the channels. In other words, the overall spectral interferogram as a function of wavelength covering the whole broadband spectrum is now divided into many parallel wavelength band channels and a combination of all the channel detector signals will produce substantially the same spectral interferogram as would have been obtained by a pure SS-OCT or SD-OCT system covering the same overall wavelength range.
As part of the calibration of the system, one can record, as a function of time in a sweep of the source, the optical frequencies received by each channel of the multi-channel receiver. In operation, the multi-channel receiver is synchronized with the multi-wavelength laser via a processor such as a personal computer (PC) and the progress of the sweep can be monitored, for example as described by Huber et al. (Optics Express 13(9): 3513-3528.). The simultaneously lasing optical frequencies will likely maintain a fixed relationship to each other, depending on the type of source. For lasers with interferometer-based filters, the simultaneously lasing optical frequencies satisfy a common interference condition, differing only in the integer order of interference. Therefore, in many cases monitoring only one of the simultaneously-lasing optical frequencies is sufficient to know the location of each frequency in the set.
The recorded partial spectral interferograms from the individual channels are combined to form the combined spectral interferogram. If the acceptance bands of each receiver channel are not strictly square pass band but with roll-off at the two sides of the pass band which is generally the case, and there is overlap between neighboring channels in the multi-channel receiver as shown in
As is well known to those skilled in the art, the speed of a pure SS-OCT system is practically limited by the sweeping time of the tunable laser to cover the whole broad spectral band rather than by the response speed of the photodetector. Due to the parallel nature of the multiple channels and the reduced tuning range for each lasing wavelength of the tunable multi-wavelength laser, the overall time required to capture the whole spectral interferogram over the whole broad spectrum range for each A-scan can thus be reduced many fold as compared to the pure SS-OCT case as long as each channel detector can respond fast enough, which is generally the case. In other words, the speed for each A-scan can be increased by a factor approximately equal to the number of lasing wavelengths.
In the preferred embodiment, the light source is a tunable multi-wavelength laser or a broadband source such as a superluminescent diode followed by a tunable comb filter that should generate at least two discrete wavelengths and the detector should have at least two channels for measuring two spectral regions. In order to achieve a reasonable benefit from the subject invention, it is desirable that light source should generate at least ten discrete wavelengths and preferably between 10 and 40 and most preferably between 10 and 20 discrete wavelengths. The detector should have a corresponding number of channels to detect a spectral region corresponding to the scan ranges for each of those discrete wavelengths. The scanning range for each wavelength should be at least 0.5 nm and is more preferably at least 1.0 nm and preferably in the range of 1.0 nm to 5.0 nm.
The particular wavelengths selected will depend upon the application. For example, in the case of ocular measurements, it is know that diagnostic OCT is typically performed in the near infrared region of the spectrum, from 700 nm to 1600 nm.
In terms of cost, the presently invented FD-OCT system will have a substantial advantage. As will be elaborated later, the proposed swept multi-wavelength laser can be fabricated using well-established technologies similar to those used for the tunable semiconductor lasers made for fiber optic telecommunication applications. The multi-channel receiver has already been developed for fiber optic communication applications and is hence commercially available at low cost.
It could perhaps also be appreciated by those skilled in the art that the pure SS-OCT configuration and the pure SD-OCT configuration are actually two extreme cases of the presently invented alternative FD-OCT configuration. In fact, when the channel number is reduced to 1, the invented alternative FD-OCT configuration is reduced to the pure SS-OCT case. On the other hand, when the channel number increases to the point (e.g. 2000 channels) that a useful spectral interferogram can be taken from the detectors without sweeping the lasing wavelengths, the invented FD-OCT system is then effectively reduced to the pure SD-OCT case. Practically speaking, there is a fundamental limit to the number of lasing wavelengths that can be generated by a simultaneous multi-wavelength laser. That limit is set by either the lasing spectral line width or mode competition as a result of homogeneous broadening of the gain optical medium of the laser. Hence there is a limit to the minimum spacing that can be achieved between two neighboring simultaneously lasing wavelengths. In theory, when the spectral resolution of the array spectrometer of a pure SD-OCT or the spectral range covered by each pixel of the line scan camera of the spectrometer is equal to the lasing line width of a multi-wavelength laser, the range required to sweep each lasing wavelength across each channel spacing is reduced to zero and the multi-wavelength laser can be viewed as effectively functioning the same as a broadband light source such as a superluminescent diode.
The multi-channel receiver is used in this invention in a way distinct from the line-scan camera in a pure SD-OCT configuration. Each channel in the multi-channel receiver records a time-dependent optical power encoding part of the overall spectral interferogram, while each pixel in a line-scan camera used for pure SD-OCT records one static power level per A-scan. The level of parallelism can be easily adjusted in the multi-channel receiver.
In terms of system-level implementation of the present invention, it should be realized that the illustrated system configuration is only a basic representation and many variations exist. For example, the optical paths can be fiber optic or bulk optic or a combination of each type. Various optical components can be included in the optical path to manipulate the beam of light. For example, one or more optical circulator(s) can be used in various parts of the interferometer to enable high optical power efficiency and also dual balanced detection as is well known to those skilled in the art (U.S. Pat. No. 6,657,727, U.S. Provisional patent application No. 60/629,492). Various optical polarization manipulation components, including polarization beam splitters, wave plates, Faraday rotators, and polarizers can all be employed for various purposes such as polarization sensitive OCT (U.S. Pat. No. 5,459,570) or polarization diversity detection (U.S. Pat. No. 5,202,745). Multiple optical phase control elements can also be used to enable multi-phase detection (US20040239938, WO04111929). In addition, an optical frequency shifter or optical phase modulator can also be used in either the sample path or the reference path to remove the mirror image and to double the depth scan range (see for example, Yun, S. H. et al. (2004) “Removing the depth-degeneracy in optical frequency domain imaging with frequency shifting.” Optics Express 12(20): 4822-4828; and Zhang, J. et al. (2005). “Removal of a mirror image and enhancement of the signal-to-noise ratio in Fourier-domain optical coherence tomography using an electro-optic phase modulator.” Optics Letters 30(2): 147-149).
The key optical components for the present invention are the swept multi-wavelength laser and the multi-channel receiver. Fortunately, in the past several years, there has been a lot of development on DWDM (dense wavelength division multiplexing) optical components for optical fiber telecommunication applications. Hence the presently invented alternative FD-OCT system can substantially take advantage of these low cost optical devices. Of particular interest are the fact that broadband semiconductor optical amplifiers and broadband discretely tunable semiconductor laser have already been developed, and the fact that multi-channel demultiplexers as well as multi-channel receivers have also already been well developed.
In particular, the technology to incorporate sampled or superstructure distributed feedback gratings directly on a semiconductor laser chip with a comb multi-wavelength reflection spectrum has been well developed (see for example, U.S. Pat. No. 4,896,325, U.S. Pat. No. 5,325,392, US20030128724, and US20040228384). While for optical fiber telecom applications, two multi-wavelength filters of different comb width are generally used as front and back mirrors to take advantage of the Vernier effect to produce discretely tunable single lasing wavelength over a large lasing bandwidth (see for example U.S. Pat. No. 4,896,325), the concept of a tunable multi-wavelength comb filter grating can be directly applied to produce swept multi-wavelength semiconductor lasers.
According to one aspect of the present invention, a tunable multi-wavelength filter 502, such as a sinc-sampled grating (see for example U.S. Pat. No. 5,325,392 and Ibsen, M. et al. (1998) “Sinc-sampled fiber Bragg gratings for identical multiple wavelength operation.” Photonics Technology Letters, IEEE 10(6): 842-844) can be made directly on or near the lasing section 504 of a multi-wavelength semiconductor laser 500 as shown in
As is well known to those skilled in the art, the maximum achievable Bragg wavelength tuning by electrical current injection into a distributed Bragg grating waveguide fabricated on a typical III-V semiconductor waveguide is about 5 to 10 nm for a near infrared center wavelength (see for example, U.S. Pat. No. 5,325,392). This tunable range is not enough for a single wavelength sweep across a broad spectral range of 25 to 200 nm but is enough to cover the channel spacing of DWDM based standard multi-channel receivers. In fact, the multi-channel receivers designed for standard DWDM ITU (International telecommunication union) grid has a channel spacing of Δf=50, 100, 200, 400, 800 GHz, which, for a center wavelength of λc=1.3 μm, corresponds to a wavelength spacing of 0.282, 0.564, 1.127, 2.254 and 4.508 nm respectively. It can thus been seen that each of the above-mentioned channel spacing is all within the tunable range that can be achieved by current-injection into a sampled grating based multi-wavelength semiconductor laser.
The corresponding sampling period P as shown in
According to another aspect of the present invention, the swept multi-wavelength laser can also be made based on a tunable Fabry-Perot (F-P) laser cavity, simply because an F-P cavity is also a comb multi-wavelength filter.
As is well known to those skilled in the art, the free spectral range or comb width of a F-P cavity is given by Δf=c/(2 neff Leff), where c is the speed of light in vacuum, neff is the effective refractive index of the F-P cavity, and Leff is the effective length of the F-P cavity. For InP with neff≈3.25, and Δf=50, 100, 200, 400, 800 GHz, the corresponding effective F-P cavity length is found to be Leff≈920 μm, 460 μm, 230 μm, 115 μm and 57.5 μm respectively. While last two cavity lengths might be too small for the desired optical output power, the tuning range required, and die handling or device packaging, the first three lengths are all reasonable for practical device fabrications.
Although semiconductor lasers with a small number of F-P cavity longitudinal modes are now commercially available (Vilokkinen, V. et al. (2004). “Reliability analysis of AlGaInAs lasers at 1.3 μm.” Electronics Letters 40(23): 1489-1490), they are single-section devices with narrow gain bandwidth. In order to use the F-P cavity approach to greater advantage of the present invention, a few adjustments to the laser can be made. The first is to substantially increase the bandwidth of the gain curve to cover the desired overall spectral bandwidth; this has already been demonstrated in recent years through, for example, quantum well intermixing (see for example, U.S. Pat. No. 6,617,188). The second improvement is to make the F-P cavity into at least two sections including a gain section and a second transparent section for refractive index modulation and hence lasing wavelength sweeping. This issue can also be addressed by quantum well intermixing (see for example, U.S. Pat. No. 6,878,562).
The third issue is the precision control of the length of the F-P cavity. This precision control may be required because the accuracy of the spacing between two neighboring lasing wavelengths depends on the accuracy with which the two facets of the cavity are formed. Fortunately, this is not a big problem for the present invention, as the multiple lasing wavelength spacing does not need to exactly match the channel width of the multi-channel receiver and can be larger than the receiver grid. Reconstruction of a full spectral interferogram from the partial spectral interferograms will succeed as long as the lasing comb spacing is equal to or greater than the channel width of the multi-channel receiver and is less than the overall sweep range.
If precision comb spacing is still preferred, lithography can be used to define the cavity length and to directly dry etch optically smooth facets to define the two end mirrors for the F-P cavity (see for example, Behfar A. et al. (2005) “Etching Advances—Replacing cleaved facets with etched facets produces improved lasers” SPIE oemagazine, February 2005, p 27).
It should be noted that the same discussion made on the F-P cavity based swept multi-wavelength laser can be easily extended to a semiconductor ring or race-track cavity. Therefore, a ring cavity based tunable multi-wavelength laser is also a possible light source for the present invention. The ring or race-track cavity based semiconductor laser preferably also has at least two sections, of which one is a gain section for lasing and the other is a transparent index modulation section for sweeping the lasing wavelengths. In addition, a separate optical amplifier can also be made next to the laser to boost the optical output power.
An issue that might need to be considered for a sampled grating or F-P cavity or ring cavity based multi-wavelength laser is the chromatic dispersion in the refractive index of the waveguide material. This dispersion will cause the lasing optical frequencies to be non-uniformly spaced if the overall optical frequency or wavelength range to be covered is large. In the present invention, the spacing of lasing frequencies does not need to match the spacing of the receiver channels, because reconstruction of the full spectral interferogram as illustrated in
According to another aspect of the invention, a tunable multi-wavelength comb filter can also be incorporated in an extended cavity of a laser. FIGS. 7 (a), (b) (c) (d) and (e) show some exemplary configurations of such a design. In these exemplary embodiments, optical fibers and other fiber optic components are preferably used to form an optical resonator with a long enough overall cavity length such that the longitudinal modes of the long cavity are so densely spaced that mode hopping between these longitudinal mode is not a concern any more for OCT applications and the actual multiple lasing wavelengths are determined by a tunable comb filter acting as part of the laser cavity.
In the first exemplary embodiment of
In the second example as shown in
In the third example as shown in
As an alternative as shown in
In addition to a ring or quasi-ring cavity, a tunable reflective multi-wavelength mirror 742 can also be used in an F-P cavity to construct a tunable multi-wavelength laser as shown in
It should be understood that we have only shown the basic optical elements for the various exemplary configurations, other optical or fiber optical components, including intensity or phase modulators, and especially polarization control components, may be included in or outside the laser cavity for optical beam or polarization manipulation. Although we have constantly mentioned the use of a semiconductor optical amplifier, this is only for the purposes of illustration as other optical amplifiers can also be used, including for example, a rare earth doped fiber amplifier. In addition, the fiber optic path can also be replaced by a free space optic path or a combination of fiber and free space optics.
It should also be understood that the basic principle of operation in terms of the tuning of the F-P cavity and the SBG is similar to what we have discussed previously, the only difference is that for piezo-electrically actuated comb filters, the change in the effective cavity length or effective grating period is caused by a physical dimensional change instead of a change in the effective refractive index. However, if an electro-optical material is used, a change in the effective refractive index can also be used.
Note that other types of multi-wavelength comb filters can all be used, including, for example, long period fiber gratings, few mode fiber Bragg gratings, Mach-Zehnder and Sagnac interferometer filters. In fact, an F-P or SFBG filters may have a limited tuning speed because they need to be mechanically stretched or compressed using for example a piezoelectric actuator. In order to achieve high tuning speed, other mechanism such as those based on electro-optical effect in optical crystals or polymers can all be used. For example, a LiNbO3 waveguide electro-optic polarization controller can be incorporated in a fiber Sagnac loop to act as a high-speed tunable multi-wavelength filter. A LiNbO3 or semiconductor based sampled grating can also be used as a high speed tunable multi-wavelength filter.
A variety of multi-channel receivers are currently available that should be adaptable for use in the subject invention. One example would include a monolithically integrated array of photodetectors 802 with a passive planar waveguide based Echelle grating 808 on a single InP chip as shown in
Another example of a receiver is shown in
Yet another option is shown in
As still another option, the multi-channel receiver can also be a combination of an independent optical demultiplexer with a number of photodetectors, as is well known to those skilled in the art of fiber optic communications. The photodetectors can be either directly arranged next to the independent demultiplexer or optically connected to the independent demultiplexer via free space or a number of optical waveguides such as optical fibers. The independent demultiplexer is preferably an arrayed waveguide grating (AWG) made on a planar lightwave circuit, or a holographic Bragg reflector grating (see for example, Greiner, C. M. et al. (2004) “Wavelength Division Multiplexing Based on Apodized Planar Holographic Bragg Reflectors” Applied Optics, Vol. 43, No. 23, pp. 4575-4583) or a planar waveguide Echelle grating (see for example, Janz, S. et al. (2004) “Planar waveguide echelle gratings in silica-on-silicon” Photonics Technology Letters, 16, pp. 503-505), or a bulk blazed diffractive grating.
Since multi-channel receivers are widely used in high speed optical fiber communication systems and are commercially available, there may not be a need for a custom made device for the present invention. On the other hand, as the channels of the multi-channel receiver generally do not have a perfect rectangular pass band spectral response, it might be necessary to make the channel slightly overlapping with each other. In addition, it may be preferred, but not required, to have the lasing frequencies spaced more widely than the receiver channels, including an allowance for the soft edges of the pass bands, in order that each receiver collects spectral interference fringes from at most one wavelength. If two different receivers partially sample the same optical frequency, these samples can be combined to form a better estimate of the interferogram at that frequency during the post-processing steps that combine the partial spectral interferogram from each individual detector.
It should be highlighted that the configuration of the present invention is relatively simple and compact. It is also of lower cost and higher speed. In addition to general optical interferometry for various applications, such as distance measurement, topography and three dimensional measurement of a volume sample, the present invention is particularly beneficial for applications in various optical coherence tomography schemes.
The foregoing description of the invention is presented for purposes of illustration and description and is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described to best explain the principles of the invention and its practical application to thereby enable others skilled in the art to best use the invention in various embodiments and with various modifications suited to the particular use contemplated.
The following patents, patent applications and other documents are hereby incorporated by reference.
U.S. Pat. No. 4,896,325
U.S. Pat. No. 5,202,745
U.S. Pat. No. 5,325,392
U.S. Pat. No. 5,459,570
U.S. Pat. No. 5,956,355
U.S. Pat. No. 6,160,826
U.S. Pat. No. 6,617,188
U.S. Pat. No. 6,657,727
U.S. Pat. No. 6,847,661
U.S. Pat. No. 6,878,562
US20030128724
US20040228384
US20040239938
US20050018201
US20050035295
U.S. Provisional patent application No. 60/629,492
CA2474331
CA2476174
DE10207186
EP1470410
EP1476716
JP2000-046729
JP2001-174404
WO03062802
WO03073041
WO04111929
WO9835203
WO2004043245
WO2005022709