Analysis of spherical aberration of a water immersion objective: application to specimens with refractive indices 1.33-1.40.

The method of using immersion medium to correct spherical aberration for water immersion objectives when the samples are not water is investigated. Spherical aberration is measured by an interferometer converted from a confocal microscope for samples with different refractive indices. When the proper refractive index of the immersion medium and thickness of cover slip are selected, the measured spherical aberration approaches zero. A theoretical model can be used for prediction of the immersion medium to correct spherical aberration for various samples. Using the thinnest available cover slip (100 microm), the zero spherical aberration condition can be applied to samples with refractive index as high as 1.40. Confocal images in the condition of almost no spherical aberration are included to demonstrate the improvement of axial resolution due to this correction.

Fiber-coupled multiplexed confocal microscope.

We describe a new parallel scanning mechanism for confocal microscopy that is inherently fiber-optic compatible and that retains the simplicity of the line scanning confocal microscope. The method works by employing an incoherent fiber-optic bundle that maps a line illumination pattern back on itself on double passing, while separating the fibers that carry photons from out-of-focus sample planes. The transformation permits efficient rejection of out-of-focus photons by a slit aperture.

Theoretical basis of confocal microscopy.

A confocal microscope forms its image by recording light primarily from a small focal volume, largely ignoring points to the side or above or below. That volume, described as a point-spread function, is the product of two similar functions that are generated by the objective lens. Because of that multiplication, the recorded light is greater than even the integrated total of the light from all other points in a thick sample. Some of the implications of implementing this theory are reflected in the choices available to users of confocal microscopes.

In vivo confocal scanning laser microscopy of human skin II: advances in instrumentation and comparison with histology.

In 1995, we reported the construction of a video-rate scanning laser confocal microscope for imaging human skin in vivo. Since then, we have improved the resolution, contrast, depth of imaging, and field of view. Confocal images of human skin are shown with experimentally measured lateral resolution 0.5-1.0 microm and axial resolution (section thickness) 3-5 microm at near-infrared wavelengths of 830 nm and 1064 nm; this resolution compares well to that of histology which is based on typically 5 microm thin sections. Imaging is possible to maximum depth of 350 microm over field of view of 160-800 microm. A mechanical skin-contact device was developed to laterally stabilize the imaging site to within +/- 25 microm in the presence of subject motion. Based on these results, we built a small, portable, and robust confocal microscope that is capable of imaging normal and abnormal skin morphology and dynamic processes in vivo, in both laboratory and clinical settings. We report advances in confocal microscope instrumentation and methods, an optimum range of parameters, improved images of normal human skin, and comparison of confocal images with histology.

Video-rate confocal scanning laser microscope for imaging human tissues in vivo.

We have built a video-rate confocal scanning laser microscope for reflectance imaging of human skin and oral mucosa in vivo. Design and imaging parameters were determined for optimum resolution and contrast. Mechanical skin-holding fixtures and oral tissue clamps were made for stable objective lens-to-tissue contact such that gross tissue motion relative to the microscope was minimized. Confocal imaging was possible to maximum depths of 350 microm in human skin and 450 microm in oral mucosa, with measured lateral resolution of 0.5-1 microm and axial resolution (section thickness) of 3-5 microm at the 1064-nm wavelength. This resolution is comparable with that of conventional microscopy of excised biopsies (histology). Normal and abnormal tissue morphology and dynamic processes were observed.

Confocal microscope with large field and working distance.

A confocal microscope that uses separate, noncollinear objective lenses for illumination and collection makes it possible to work 20 mm from a sample with a field of view of 2.25 mm across and still achieve lateral and axial resolutions below 5 microm. The design is expected to have application to in vivo confocal microscopy, allowing a field of view and section thickness similar to those used by pathologists in examining excised tissue.

Measurement of the wave-front aberration of the eye by a fast psychophysical procedure.

We used a fast psychophysical procedure to determine the wave-front aberrations of the human eye in vivo. We measured the angular deviation of light rays entering the eye at different pupillary locations by aligning an image of a point source entering the pupil at different locations to the image of a fixation cross entering the pupil at a fixed location. We fitted the data to a Zernike series to reconstruct the wave-front aberrations of the pupil. With this technique the repeatability of the measurement of the individual coefficients was 0.019 micron. The standard deviation of the overall wave-height estimation across the pupil is less than 0.3 micron. Since this technique does not require the administration of pharmacological agents to dilate the pupil, we were able to measure the changes in the aberrations of the eye during accommodation. We found that administration of even a mild dilating agent causes a change in the aberration structure of the eye.

Spectrally encoded confocal microscopy.

An endoscope-compatible, submicrometer-resolution scanning confocal microscopy imaging system is presented. This approach, spectrally encoded confocal microscopy (SECM), uses a quasi-monochromatic light source and a transmission diffraction grating to detect the reflectivity simultaneously at multiple points along a transverse line within the sample. Since this method does not require fast spatial scanning within the probe, the equipment can be miniaturized and incorporated into a catheter or endoscope. Confocal images of an electron microscope grid were acquired with SECM to demonstrate the feasibility of this technique.

In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast.

Confocal scanning laser microscopy of live human skin was performed to investigate the correlation of in vivo cellular and morphologic features to histology, the effect of wavelength on imaging, and the role of melanin as a contrast agent. We built a video-rate confocal scanning laser microscope for in vivo imaging of human skin. Using a 100 x microscope objective, we imaged high-contrast optical "sections" of normal skin, vitiliginous skin, and a compound nevus. In vivo "confocal histology" correlated well with conventional histology. The maximum imaging depth increased with wavelength: the epidermis was imaged with visible 400-700-nm wavelengths; the superficial papillary dermis and blood cells (erythrocytes and leukocytes) in the deeper capillaries were imaged with the near infrared 800-900-nm wavelengths. For confocal reflectance imaging, melanin provided strong contrast by increased backscattering of light such that the cytoplasm in heavily pigmented cells imaged brightly. In vivo confocal microscopy potentially offers dermatologists a diagnostic tool that is instant and entirely non-invasive compared to conventional histopathology.

Dynamics of external ocular blood flow studied by scanning angiographic microscopy.

The scanning angiographic microscope (SAM) provides a solution to the considerable technical difficulties associated with conventional episcleral fluorescein angiography. Standardised anterior segment fluorescein videoangiograms were performed using the SAM in each episcleral quadrant of the right eye in 6 normal subjects; frame-by-frame analysis proved important. Centripetal flow was seen in all 37 scleral perforating arteries investigated. Other features were the marked individual variability, much larger vertical anterior ciliary arteries, the high frequency of arteriovenous anastomoses, the complex flow patterns, the absence of a 'watershed' zone between anterior ciliary and posterior episcleral circulations, a characteristic and discontinuous distribution of 'leaky' episcleral veins, and the primacy of venous drainage into the plexus of muscular veins. Reports of retrograde blood flow in the anterior ciliary arteries in most fluorescein angiographic studies are probably incorrect, the result of unappreciated methodological problems. The SAM is an important advance on previous anterior segment fluorescein angiography techniques.

Microlaser microscope using self-detection for confocality.

The microlaser microscope can use its own source lasers as detectors to provide a matched array of confocal apertures. Detection of the laser light by a single avalanche photodiode makes the electronics simple. We have implemented this in a 64 x 312 pixel format as our first demonstration of the device.

Plate beamsplitter to produce multiple equal-intensity beams.

The design of a plate beamsplitter to produce multiple beams of equal intensity is presented. Multiple beams of equal intensity can be obtained from a plate by varying the reflectances of the front and back surfaces. The application for which we designed the plate beamsplitter was a fourbeam multiplexed galvanometric scanner for a confocal scanning microscope. Multiplexing with four beams increases the effective optical scanning rate (and therefore the confocal imaging rate) to four times the electromechanical scanning rate of the galvanometrically driven mirrors.

Mapping cone photopigment optical density.

The distribution of cone photopigment across the retina affects the amount of light captured by cones at each retinal location. Cone photopigment optical density is measured in two ways, with reflectometry and/or with color matching. Color matching measures a higher optical density than does reflectometry. Control experiments confirm that large-field color matches measure photopigment optical density toward their outer edge. There is qualitative agreement as to photopigment distribution from both techniques near the fovea. Beyond 1 deg, color matching indicates little decrease in photopigment with increasing eccentricity, whereas retinal densitometry shows a steep decline in photopigment. The decrease in perifoveal optical density measured with reflectometry is attributed to the decrease in cone coverage from fovea to perifovea as rods and interphotoreceptor spaces increase. Differences among subjects in photopigment distribution near the fovea, measured with both techniques, reflect differences in the specialization of the foveal center for cone length and/or photopigment concentration per cone, which are factors influencing results from both techniques.

Detectors for scanning video imagers.

Detectors for scanning video (10-MHz) imagers should be chosen for their high quantum efficiency and internal gain. Because of the high bandwidth both photomultiplier tubes and avalanche photodiodes are limited by photon noise, so that dark noise is not the determining quantity.

Measurement of ocular local wavefront distortion with a spatially resolved refractometer.

Local wavefront distortion by the total refractive system of the eye is measured by a variant of the Scheiner principle at some thirty loci of 1 mm diameter. At each locus we find the normal to the wavefront that could form a point focus. A simple visual display is used to review the data, and a steepest descents fit of the wavefronts with a power series enables comparison with more traditional measures of refractive error.

Reflectometry with a scanning laser ophthalmoscope.

We describe noninvasive techniques to optimize reflectometry measurements, particularly retinal densitometry, which measures the photopigment density difference. With these techniques unwanted scattered light is greatly reduced, and the retina is visualized during measurements. Thus results may be compared for each retinal location, and visible artifacts are minimized. The density difference measurements of the cone photopigment depend on the optical configuration of the apparatus. The cone photopigment density difference is greatest near the fovea and for most observers decreases rapidly with eccentricity. A research version for reflectometry and psychophysics of the scanning laser ophthalmoscope is described.

Anterior segment fluorescein videoangiography with a scanning angiographic microscope.

The scanning laser ophthalmoscope can be modified to operate as a scanning laser biomicroscope for use in anterior segment fluorescein angiography. The substantial depth of focus, large field of view, co-axial illumination, low light levels, real-time television operation, and videorecording with immediate recall provide advantages not available with conventional photographic methods. Video techniques give a resolution slightly inferior to photography, but this is unlikely to be significant in clinical practice. A technique of traversing the entire anterior episcleral vasculature has been developed to give a comprehensive and reproducible angiographic record. Previous fluorescein studies suggesting the primary importance of retrograde (centrifugal) flow in the perforating anterior ciliary arteries were not supported; methodologic explanations are advanced. Several principles are proposed to improve techniques of anterior segment angiography.

Confocal scanning laser ophthalmoscope.

A confocal scanning imager moves an illumination spot over the object and a (virtual) detector synchronously over the image. In the confocal scanning laser ophthalmoscope this is accomplished by reusing the source optics for detection. The common optical elements are all mirrors-either flat or spherical-and the scanners are positioned to compensate astigmatism due to mirror tilt. The source beam aperture at the horizontal scanner is small. Light returning from the eye is processed by the same elements, but now the polygon's facet is overfilled. A solid-state detector may be at either a pupillary or retinal conjugate plane in the descanned beam and still have proper throughput matching. Our 1-mm avalanche photodiode at a pupillary plane is preceded by interchangeable stops at an image (retinal) plane. Not only can we reject scattered light to a degree unusual for viewing the retina, but we choose selectively among direct and scattered components of the light returning from the eye. One (of many) consequences is that this ophthalmoscope gives crisp and complete retinal images in He-Ne light without dilation of the pupil.