| Literature DB >> 7265186 |
Abstract
Bronchial biopsies from ten subjects, including five smokers, have been examined using light microscopic morphometry. The biopsies were free from identifiable disease. Using manual point counting and a Quantimet 720 image analysing computer, a number of parameters were measured. Computer-based cluster analysis of seven of these parameters associated the subjects into three groups: visual inspection of the sections achieved the same separation. Four subjects (non-smokers) had normal epithelia: four subjects (one non-smoker, three smokers) showed mucous cell hyperplasia: two subjects (smokers) had squamous cell metaplasia. Three parameters in conjunction contained sufficient information to characterize accurately the histological appearance of the epithelia: the epithelial thickness, the volume density of intracellular mucus and the number of nuclear profiles per unit area of sectioned epithelium. Reduction of these three parameters to a linear plot closely approximated a similar reduction of the original seven parameters. These three parameters can be measured rapidly, either manually or by the Quantimet. The linear representation of these parameters provides a reproducible and objective basis for comparing specimens of bronchial epithelium.Entities:
Mesh:
Year: 1981 PMID: 7265186 DOI: 10.1111/j.1365-2818.1981.tb01280.x
Source DB: PubMed Journal: J Microsc ISSN: 0022-2720 Impact factor: 1.758