Marvin Minsky invented the confocal scanning microscope while working as a postdoctoral fellow at Harvard in 1955.9 However, its employment to image tissue in vivo required extensive development of light sources and computerization technologies. Since the 1980s, several research groups have employed tandem scanning confocal microscopy to image human and animal tissue in vivo.10–13 The first reported use of confocal scanning laser microscopy to image human skin in vivo dates back to 1995.8 Since then, many reports have suggested its possible applications in dermatology, particularly in the diagnosis of skin tumors.

In vivo RCM offers several important advantages over conventional histology. Imaging is painless and non-invasive, causing no tissue damage. The skin is not altered by post-collection processing (fixation, sectioning, and mounting) or staining, minimizing artifacts or disruption of the native structure of the tissue. Real-time data collection is faster than routine histology, and the same location of the skin can be repeatedly imaged over time to evaluate dynamic changes such as tissue growth, wound healing, lesion progression or response to therapy.14–18 This is one of the main advantages of the technique compared to conventional histology, which can only reveal information about the tissue at the time of collection of the sample.

Reflectance confocal microscopy is based on the use of a point source of light that illuminates a small spot within translucent tissue, such as skin or mucosa. The reflected light (reflectance) is imaged onto a detector after passing through a small pinhole. This pinhole prevents out-of-focus light from reaching the detector, so that only light from the exact in-focus plane (confocal) is detected. An entire image of the whole plane of the sample under study is generated using line scanning of the point source beam. This enables a virtual sectioning of a thin horizontal tissue plane in vivo, such as computerized tomography or MRI. Such optical sectioning of the tissue allows intact tissue to be studied without physical sectioning, as in conventional histology, preserving the physical structure and physiology of the tissue.

The resolution depends on the pinhole size, the numerical aperture of the objective lens, and the wavelength used. Although lasers of different wavelengths may be used as the light source for RCM, only longer wavelengths penetrate deep enough into the skin. Therefore, confocal microscopes use sources of illumination near the infrared wavelength (800–1,064 nm). Backscattering of light occurs due to local variations of the refractive index within the tissue as well as when the scattering structure has a size similar to the illuminating wavelength. For example, near-infrared wavelengths produce strong backscattering from melanosomes (despite melanin absorption at this wavelength) because they contain material with a high refractive index relative to the surrounding skin structures and have a size similar to the illuminating wavelength.14 This can be used in order to detect cells containing melanin, such as basal keratinocytes, melanocytes, and melanophages, which appear very bright upon imaging. Other microstructures similar in size to the illuminating wavelength, such as organelles or granules in the cytoplasm also act like reflectance surfaces and may determine other endogenous sources of contrast in Langerhans cells and lymphocytes.

New microscopes use a single optical fibre that is common to both the illumination optical path and the detection optical path. This single-moded optical fibre acts as a spatial filter that discards out-of-focus information collected by the objective lens. This system simplifies the opto-mechanical complexity by eliminating the bulky optical components and confocal apertures (pinholes) associated with conventional confocal implementations. This fiber-optic approach has enabled miniaturization of the confocal scanner into a hand-held device that provides the flexibility and mobility necessary for actual clinical use.

The currently commercially available RCM has a wavelength of 830 nm and 30× objective lens of NA 0.9, which provides a lateral resolution of approximately 1 μm and an axial resolution (section thickness) of 3–5 μm.14 It is possible to image normal skin from the stratum corneum to a depth of 200–250 μm, at the level of upper reticular dermis.14 Imaging deeper layers within the skin may also be achieved by using greater laser power, but the laser power used in the commercially available device is less than 30 mW because it causes no tissue damage or eye injury. Water immersion lenses are used since the refractive index of water (1.33) is close to that of the epidermis (1.34), which minimizes spherical aberrations caused when the light passes through the tissue-air interface.14 Other immersion media, such as water-based gels, are useful when imaging a scaly or hyperkeratotic lesion since the gel settles between disrupted corneocytes, reducing refractive irregularities. Such gels are also useful when imaging irregular skin sites, since the gel does not run off of the skin the way water or other aqueous solutions would do.

Skin movement is an important limitation when imaging in vivo. To reduce motion artefacts and contain the water or the gel interface when imaging, a skin contact device is used. This device consists of a metal ring that is fixed to the skin with adhesive, and is coupled to the microscope housing with a magnet during imaging. It has a concave shape to hold the immersion medium over a round window, which allows passing of light on to the sample as well as reflected light from the tissue.

Images are obtained after sequential illumination of multiple pixels according to the density of the reflected light from focal plane. This plane is scanned in two dimensions in order to create a mosaic image; architectural patterns can be evaluated this way. By moving the objective lens in the “z” stage (vertical) with respect to the skin surface, it is possible to image at different horizontal levels within the tissue since the focal plane is progressively moved deeper or closer to the surface. We can select each imaged square to compose a mosaic image and study cytological or architectural details. Images can be captured to produce static pictures of horizontal skin sections as well as recorded on videotape (20–30 frames per second) to produce movies that reveal dynamics events such as blood flow.8,14–18

Real-time reflectance-mode confocal microscopy offers a novel view of the skin in terms of the orientation and image content. Compared to routine histology, it offers two main differences. First, the image obtained is horizontal or en face rather compared to the vertical sections typical in histology. Second, images are greyscale (bright scale), similar to radiographs or echographic examinations. The field of view provided by RCM varies with different microscopes, but is generally 250–500 μm across.14 The layer of the skin under study can be ascertained by the morphologic appearance of tissue at a given depth or by measuring the depth of section, using a micrometer attached to the “z” stage of the objective lens.

When imaging the skin, the most superficial images obtained are of the stratum corneum. The morphologic appearance of that section includes anucleated polygonal corneocytes 10–30 μm in size grouped in “islands” separated by skin folds, which appear very dark (Fig. 1a). Images are very bright because the refractive difference at the interface between the immersion medium and stratum corneum results in a large amount of back-scattered light. The next layer is the stratum granulosum, which consists of two to four layers of cells measuring 25–35 μm. The nuclei appear as dark central ovals within the cell surrounded by a bright grainy cytoplasm (Fig. 1b). The next layer is the stratum spinosum, which consists of a tight honeycomb pattern of smaller cells 15–25 μm in size with well-demarcated cell borders (Fig. 1c). The deepest layer of the epidermis is the basal layer, visualized as bright clusters of cells due to their content in melanin, 7–10 μm in size. The suprapapillary epidermal plate at the dermo-epidermal junction appears as rings of bright basal cells surrounding dark dermal papillae, which often show a central area of blood flow consistent with papillary dermal vascular loops (Fig. 1d). The papillary dermis images consist of a network of reticulated fibres with central capillary vessels (Fig. 1e). In the reticular dermis, confocal images show a thick reticulated pattern consisted of collagen bundles and elastic fibres (Fig. 1f). We can also observe eccrine ducts, which appear as bright, centrally hollow structures that spiral through epidermis and dermis, and hair shafts with pilo-sebaceous units. These appear as whorled, centrally hollow structures with elliptical, elongated cells at the circumference and a central, refractile long hair shaft.

The appearance of normal skin varies according to the anatomical site, sun-exposure, skin color, age, and physiological condition being imaged.19 Skin from sun-exposed or darkly pigmented sites, for example, appears generally brighter because of what appears to be more pigment at the basal layer.

The characterization of skin neoplasm using RCM is an important area of clinical research with potential for non-invasive diagnosis and management of a variety of skin cancers. The advent of newer, less invasive topical therapies such as topical imiquimod or photodynamic therapy demands the development of non-invasive diagnostic tools to identify tumor subtypes and tumor margins accurately, and to monitor the response to treatment.