This paper addresses fundamental issues that underlie the interpretation of images acquired from turbid tissues by optical-coherence tomography (OCT). The attenuation and backscattering properties of freshly excised rat arteries and their dependence on the focusing and collection optics of the OCT system were measured at two wavelengths in the near infrared (830 nm and 1300 nm). Determined from the ratio of the magnitudes of the reflections from glass plates placed on both sides of the arteries, the mean attenuation coefficient of the arterial wall was found to be in the range 14 < microt < 22 mm(-1) at 830 nm and 11 < microt < 20 mm(-1) at 1300 nm. The measured values of microt were lowest for the longer source wavelength and for probe beams with the smallest average diameters. The observed dependence of microt on beam size indicates that relatively large-scale variations in the index of refraction of the tissue contributed to degradation of the tranverse spatial coherence of the beam. We introduce a framework for understanding and quantifying beam-size effects by way of the mutual-coherence function. The fact that spatial variations in backscattering and attenuation (which includes spatial-coherence losses) have similar effects on OCT signals makes the origin of the signals difficult to determine. Evidence is given that suggests that, in spite of this difficulty, certain features of microstructures embedded several hundred micrometres deep in a turbid tissue can still be detected and characterized.

Background: A new type of microscope has been developed for acquiring cross-sectional images of living skin noninvasively. It takes advantage of the short temporal coherence of a broad-band light source to reject scattered light. Because this microscope is still in an early stage of development, its potential as a diagnostic tool in dermatology has not yet been determined. Objective: This study was designed to explore potential applications of optical coherence microscopy in dermatology. The aim was to investigate the structures in skin that can be seen without staining or using sophisticated image-processing methods. Methods: A prototype fiberoptic microscope was assembled that uses a 1,300-nm light-emitting diode as a light source. Scans were obtained from the skin on the index finger and forearm. Subsurface structures were identified based on knowledge of the anatomy of normal healthy skin. Results: Structures located as deep as 1 mm below the surface of the skin could be imaged with a resolution of about 10 μm in the axial and lateral dimensions. In optical slices taken perpendicular to the skin surface, the contours of the epidermal ridges and the boundary between the epidermis and dermis were readily observed. Conclusions: The results of this study suggest that an optical coherence microscope may have value as a diagnostic tool for cases in which visualization of subcellular details is not required. The resolution, contrast and scanning speed of the microscope need to be improved.

We show that the multiple-scatter rejection provided by optical coherence microscopy (low-coherence interferometry) can be incomplete in optically turbid media and that multiple scattering manifests itself in two distinct ways. Multiple small-angle scattering results in an effective probe field that is stronger than expected from a first-order beam extinction model, but that contains a distorted wave front that enhances the apparent reflectance of small structures relative to those that are larger than the unscattered incident beam. Multiple wide-angle scattering produces a broad diffuse haze that reduces the contrast of subsequent features.

We examine the performance of confocal microscopes designed for probing structures embedded in turbid media. A heuristic scheme is described that combines a numerical Monte Carlo simulation of photon transport in a turbid medium with a geometrical ray trace through the confocal optics. To show the effects of multiple scattering on depth discrimination, we compare results from the Monte Carlo simulations and scalar diffraction theory. Experimental results showing the effects of the pinhole diameter and other variables on imaging performance at various optical depths in suspensions of polystyrene microspheres were found to correspond well with the Monte Carlo simulations. The major conclusion of the paper is that the trade-off between signal level and background scattered-light rejection places a fundamental limit on the sectioning capability of the microscope.

No abstract