Current-generation adaptive optics scanning laser ophthalmoscopes (AOSLO) are typically able to compensate for the optical aberrations of the human eye and achieve diffraction-limited illumination over a wide vergence range. However, to maximize the light collection efficiency of such a system, it is also necessary to have a collection path that is diffraction-limited for each imaging channel. Although it is trivial to achieve a high collection efficiency for a single channel AOSLO or a multi-channel AOSLO with minimal vergence differences between channels, for larger vergence differences, a careful consideration of the collection optics and their position and orientation within the collection path is needed. We present a methodology and system design for achieving diffraction-limited performance in both the illumination and collection paths of a multi-color AOSLO. The system consists of three imaging channels spanning different wavelength ranges (543 ± 11 nm, 680 ± 11 nm, and 840 ± 6 nm, respectively) and one near-infrared wavefront sensing channel (940 ± 5 nm). The maximum vergence difference between channels (measured at the exit pupil of the system) is ~1.2 diopters to compensate for the chromatic focal shift of the human eye. Preliminary imaging results from a healthy human adult volunteer demonstrate the system’s ability to resolve the foveal cone mosaic in all three imaging channels.
Adaptive Optics Scanning Laser Ophthalmoscopy (AOSLO) is a powerful tool for retinal imaging at a cellular level. In conventional AOSLO, a pinhole conjugate to the retinal plane is used to reject out-of-focus light. This pinhole enables resolution improvement compared to standard bright-field microscopy imaging; however, the maximum resolution improvement can only be achieved when the pinhole is very small compared to the system’s point spread function. Thus, in practice one must find a compromise between the system’s achievable resolution and light throughput, which dictates the signal-to-noise ratio (SNR). In retinal imaging SNR is of high importance since the amount of light is limited by safety regulations.
We introduce a new detection scheme into AOSLO using a similar technique to that recently developed for confocal scanning microscopy. Here we replace the single element pinhole detector with a multi-element design. This multi-element detector consists of a multimode fiber bundle with seven fibers, where the fiber tips are arranged in a closely-spaced hexagonal pattern on one side and are completely separated on the other side and each fiber is attached to a detector. Each of the fibers acts as a pinhole recording an image which is shifted with respect to the center fiber’s image. By aligning the images of the individual fibers and consolidating all of the light collected, an enhanced image with improved SNR can be processed.
Thus, this design enables better SNR, without sacrificing resolution or altering imaging conditions. To demonstrate the capabilities of our system, we present phantom sample images.
We present a simple technique which uses a random phase object for single-shot characterization of an optical system's phase transfer function. Existing methods for aberration measurement typically involve holography, requiring complicated wavefront sensing optics or through-focus measurements with known test objects (e.g. pinholes, fluorescent beads) for pupil recovery from the measured wavefront. Here, it is demonstrated that a weak diffuser can be used to recover the pupil of an imaging system in a single measurement, without exact knowledge of the diffuser's surface. Due to its stochastic nature, the diffuser scatters light to a wide range of spatial frequencies, thus probing the entire pupil plane. A linear theory based on the weak object approximations predicts the spectrum of the measured speckle intensity to depend directly on the pupil function. Numerical simulations of diffusers with varying strength confirm the validity of the theory and indicate sufficient conditions under which diffusers act as weak phase objects. Using index matching oils to modulate diffuser strength, experiments are shown to successfully recover aberrations from an optical system using coherent illumination. Additionally, this technique is applied to the recovery of defocus in images of a weak phase object obtained through a commercial microscope under partially coherent illumination.
Non-contact imaging methods to distinguish between healthy tissue and brain tumor tissue during surgery would be highly desirable but are not yet available. Optical Coherence Tomography (OCT) is a non-invasive imaging technology with a resolution around 1-15 μm and a penetration depth of 1-2 mm that may satisfy the demands. To analyze its potential, we measured ex vivo human brain tumor tissue samples from 10 patients with a Spectral Domain OCT system (Thorlabs Callisto: center wavelength of 930 nm) and compared the results with standard histology. In detail, three different measurements were made for each sample. First the sample was measured directly after surgery. Then it was embedded in paraffin (also H and E staining) and examined for the second time. At last, the slices of each paraffin block cut by the pathology were measured. Each time a B-scan was created and for a better comparison with the histology a 3D image was generated, in order to get the corresponding en face images. In both, histopathological diagnosis and the analysis of the OCT images, different types of brain tumor showed difference in structure. This has been affirmed by two blinded investigators. Nevertheless the difference between two images of samples taken directly after surgery is less distinct. To enhance the contrast in the images further, we employ Spectroscopic OCT and pattern recognition algorithms and compare these results to the histopathological standard.
When light interacts with a scattering medium, the spectrum of the incident light undergoes changes that are dependent on the size of the scatterers in the medium. Spectroscopic Optical Coherence Tomography (S-OCT) is a method that can be used to ascertain the resulting spatially-dependent spectral information. In fact, S-OCT is sensitive to structures that are below the spatial resolution of the system, making S-OCT a promising tool for diagnosing many diseases and biological processes that change tissue structure, like cancer. The most important signal processing steps for S-OCT are the depth-resolved spectral analysis and the calculation of a spectroscopic metric. While the former calculates the spectra from the raw OCT data, the latter analyzes the information content of the processed depth-resolved spectra. We combine the Dual Window spectral analysis with different spectroscopic metrics, which are used as an input to colorize intensity based images. These metrics include the spectral center of mass method, principal component (PCA) and phasor analysis. To compare the performance of the metrics in a quantitative manner, we use a cluster algorithm to calculate efficiencies for all methods. For this purpose we use phantom samples which contain areas of microspheres of different sizes. Our results demonstrate that PCA and phasor analysis have the highest efficiencies, and can clearly separate these areas. Finally we will present data from cartilage tissue under static load in vitro. These preliminary results show that S-OCT can generate additional contrast in biological tissue in comparison to the pure intensity based images.
Digital holography (DH) is capable of providing three-dimensional topological surface profiles with axial resolutions in the nanometer range. To achieve such high resolutions requires an analysis of the phase information of the reflected light by means of numerical reconstruction methods. Unfortunately, the phase analysis of structures located in scattering media is usually disturbed by interference with reflected light from different depths. In contrast, low-coherence interferometry and optical coherence tomography (OCT) use broadband light sources to investigate the sample with a coherence gate providing tomographic measurements in scattering samples with a poorer depth-resolution of a few micrometers. We propose a new approach that allows recovering the phase information even through scattering media. The approach combines both techniques by creating synthesized interference patterns from scanned spectra. After applying an inverse Fourier transform to each spectrum, we yield three-dimensional depth-resolved images. Subsequently, contributions of photons scattered from unwanted regions are suppressed by depth-filtering. The back-transformed data can be considered as multiple synthesized holograms and the corresponding phase information can be extracted directly from the depthfiltered spectra. We used this approach to record and reconstruct holograms of a reflective surface through a scattering layer. Our results demonstrate a proof-of-principle, as the quantitative phase-profile could be recovered and effectively separated from scattering influences. Moreover, additional processing steps could pave the way to further applications, i.e. spectroscopic analysis.
Spectroscopic optical coherence tomography (OCT) is an extension of the standard backscattering intensity analysis of
OCT. It enables depth resolved monitoring of molecular and structural differences of tissue. One drawback of most
methods to calculate the spectroscopic data is the long processing time. Also systematic and stochastic errors make the
interpretation of the results challenging. Our approach combines modern signal processing tools with powerful graphics processing unit (GPU) programming. The processing speed for the spectroscopic analysis is nearly 3 mega voxel per second. This allows us to analyze multiple B-Scans in a few seconds and to display the results as a three dimensional data set. Our algorithm contains the following steps in addition to the conventional processing for frequency domain OCT: a quality map to exclude noisy parts of the data, spectral analysis by short time Fourier transform, feature reduction by Principal Component Analysis, unsupervised pattern recognition with K-means and rendering of the gray scale backscattering OCT data which is superimposed with a color map that is based on the results of the pattern recognition algorithm. Our set up provides a spectral range from 650-950nm and is optimized to suppress chromatic errors. In a proof-of-principle attempt, we already achieved additional spectroscopic contrast in phantom samples including scattering microspheres of different sizes and ex vivo biological tissue. This is an important step towards a system for real time spectral analysis of OCT data, which would be a powerful diagnosis tool for many diseases e.g. cancer detection at an early stage.
We use Spectroscopic Optical Coherence Tomography (S-OCT) to identify substances by their spectral features in multi
layer non-scattering samples. Depth resolved spectra are calculated by a windowed Fourier Transform in the spatial
regime at discrete layer borders. By dividing subsequent spectra in an iterative manner transfer functions of the samples
layers are calculated. Estimating these spectral transfer functions with high accuracy is still challenging, since the
system´s transfer function introduces an error, which can be orders of magnitude higher than the spectroscopic
information of the sample. We retrieve the buried spectroscopic information of the sample with high accuracy by
correcting the spectral transfer functions with an identically structured reference sample. This spectral calibration method
has many critical parameters and is in many cases not even possible. To perform substance identification without spectral
calibration we implemented a pattern recognition algorithm, which allocates the transfer functions to known substances.
Our results show that substance identification by spectral features with high performance without spectral calibration is
feasible. Aside from that we modeled a simplified set up of our OCT system to minimize the error which is introduced
by the optical system. The error can be reduced by orders of magnitude, when our improved optical set-up is used. This
is an important step towards an improved system for S-OCT.
We present a method to obtain additional depth resolved spectroscopic information from standard
frequency domain optical coherence tomography (FDOCT) images. This method utilizes Fourier transforms of signal peaks within the complex FDOCT depth profiles to extract depth resolved spectroscopic information. For verification of the depth resolved spectroscopic image analysis method, theoretical simulations as well as experimental studies are demonstrated. Both show accurate depth resolved spectroscopic reconstruction enabling a depth allocation of material specific transmission spectra due to absorption. This analysis tool improves significantly the image contrast and allows image mapping of material specific spectral characteristics.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.