Histol Histopathol (1998) 13: 921-926 001: 10.14670/HH-13.921 http://www.hh.um.es Histology and Histopathology From Cell Biology to Tissue Engineering Invited Review Spectral imaging for quantitative histology and cytogenetics C. Rothmannl, I. Bar-Am 2 and Z. Mallk1 1 Life Sciences Department, Bar lIan University and 2Applied Spectral Imaging. Migdal Haemek, Israel Summary. Evaluation of cell morphology by bright field microscopy is the pillar of histopathological diagnosis. The need for quantitative and objective parameters for diagnosis gave ri se to the development of morphometric methods. Morphometry combined with spectral imaging provides multi-pixel information from a specimen, which can be used for further image proces sing and quantitative analysis. The spectroscopic analysis is based on the ability of a stained histological specimen to absorb, reflect, or emit photons in ways characteristic to its interactions with specific dyes. Spectral information obtained from a histological specimen is stored in a cube whose appellate signifies the two spatial dimensions of a flat sample (x and y) and the third dimension, the spectrum , representing the light intensity for every wavelength. By mathematical analysis of the cube data- base, it is possible to perform the function of spectral- similarity mapping (SSM) which enables the demarcation of areas occupied by the same type of material. Spectral similarity mapping constructs new images of the specimen , revealing area s with similar stain-macromolecule characteristics and enhancing subcellular features. Spectral imaging combined with SSM reveals nuclear organization and identifies specifically the nucleoli domains. Therefore, differentiation stages as well as apoptotic and necrotic conditions are easily quantified . The commercial Spectra Cube TM system was developed for the application of spectral imaging in biology, recording both transmitted light and fluorescence. The SKyTrvt technique utilizes the advantages of the Spectra Cube ™ for multi probe FISH and chromosome karyotyping, identifying marker chromosomes , detecting subtle chromosome trans locations and clarifying complex karyotypes. Key words: Spectral-imaging , Fourier spectroscopy, Spectral similarity mapping, Optical density, SKY Offprint requests to: Prof. Zvi Malik. Life Sciences Department. Bar /Ian University. Ramat·Gan 52900 , Israel. Fax: 972-3-5345878. e-mail: malikz@mail.biu.ac.il 1. Principles of spectral Imaging The histopathological examination experienced an important technical improvement with the development of image analysis techniques. Histopathological classification and malignancy grading of tumors have traditionally been based on subjective qualitative evaluation of morphology seen in two-dimensional tissue sections (Sorensen, 1992). However, as Mather (1953) has pointed out : "no science could reach its flowering until it becomes quantitative in both its observation and its theory". Thus, morphometrical studies have been recommended in order to achieve an objective diagnosis (Weibel and Elias, 1967). The development of spectroscopic analysis for histological and cytological specimens provides an additional tool for objective and quantitative diagnosis. Spectroscopic analysis is based on the ability of histological and cytological specimens to absorb, reflect, or emit photons in ways characteristic to their inter- actions with specific stains. A high-resolution trace of the intensity of light radiation versus wavelength forms a graphical record unique to a given material. These characteristic absorption and emission bands occur in narrow wavelength ranges, 10 nm or less, and unless the measuring instrument has that resolution, these details cannot be detected. Spectral imaging, as the terminology suggests, combine s spectroscopy and imaging. In dramatic contrast to conventional microscopy in which fluorochrome discrimination is based on the measure- ment of a single intensity through a specific optical filter, spectral imaging allows one to measure and analyze the full spectrum of light at all pixels of an image. This technology is highly data intensive, requiring the collection of hundreds of spectral data points for every pixel of an image. Other methods utilize limited spectral data obtained from single points or lines along the cell, so that image reconstruction on the basis of the measured spectra is necessary. Spectral image analysis creates a unique data-base which enables demarcation of features and evaluation of quantities from mUlti-point spectral information of an histological specimen. The correlation between spectrum and composition is easy in the study of known molecules; in