This study describes a technique for noninvasive determination of the surface area and volume of chondrocytes using the confocal scanning laser microscope, and the fundamental limitations associated with its application. Using geometric modelling principles, an isointensity surface contour was formed from a series of optical sections recorded with the confocal microscope. Using a combined surface- and volume-based algorithm, the surface area, volume and other morphometric descriptions were calculated from a polygonal description of the cell surface. The high image contrast required for repeatable identification of the cell border was achieved through the use of a fluorescent dye, which was excluded from cells by an intact membrane. Calibration results indicated that the theoretical modelling algorithm is relatively precise when applied to simulated convex (ellipsoidal) cells, with overall errors of less than 0·5% in surface area and volume measurements. When applied to low-noise, high-contrast volume data recorded on the confocal microscope, typical coefficients of variation of 2–4% were determined for length measurements, 2–5% for volume measurements and 3–6% for surface area measurements either for latex microspheres or for chondrocytes. While the precision of the method is comparable to standard histological techniques, its accuracy is difficult to assess, as systematic errors are unpredictable and may be introduced from several sources.
Keywords:
Confocal microscopy; image analysis; image processing; three-dimensional reconstruction; morphometry; stereology; fluorescent indicator; cartilage; chondrocyte; mechanotransduction; extracellular matrix