A COMPUTER-SIMULATION STUDY ON BIOAEROSOL COLONY COUNTING ERROR DUE TO MASKING EFFECT

Citation
Cc. Chen et al., A COMPUTER-SIMULATION STUDY ON BIOAEROSOL COLONY COUNTING ERROR DUE TO MASKING EFFECT, The Annals of occupational hygiene, 42(7), 1998, pp. 501-510
Citations number
15
Categorie Soggetti
Toxicology,"Public, Environmental & Occupation Heath
ISSN journal
00034878
Volume
42
Issue
7
Year of publication
1998
Pages
501 - 510
Database
ISI
SICI code
0003-4878(1998)42:7<501:ACSOBC>2.0.ZU;2-M
Abstract
The prevalence of bioaerosol particles has gained significant attentio n in the past decades and is considered to be a potential health hazar d for people in a variety of environments, To identify and quantify ai rborne microorganisms, the bioaerosols are collected directly on semis olid culture media, or trapped from air into liquid substrates, follow ed by culturing, counting, and identification of the resulting colony forming units, Studies on error due to overcrowding of the colonies ha ve concluded that the bias did not become significant until about 300 colonies/plate were present, but no study has taken the non-homogeneit y of colony size into account, In the present study, a Cartesian coord inate system was used to simulate the surface area of standard 100 mm diameter colony culture plates and the colonies growing on them, A res olution of 1000 x 1000 was found to be accurate enough to represent th e actual size of the colony, The capacity of distinguishing overlappin g colonies (defined as resolution index) the covering fraction, the co lony surface density, and the colony size distribution were among the primary parameters investigated in this study. The counting efficiency was found to consistently decrease with increasing surface density an d decreasing resolution index, The recommended upper Limit of colony c ounts on each plate depended not only on the number concentration but also the size distribution of the colonies. To have a minimum counting efficiency of 95%, the colony densities should not exceed 1290, 200 a nd 71 colonies/plate, for colony sizes of 2, 5 and 8 mm, respectively. This was based on the assumption that the colony size was uniform and resolution index was 0.8. As the geometric standard deviation (GSD) k ept increasing, the Larger GSDs resulted in higher covering fraction, and thus, lower counting efficiency. This phenomenon was independent o f resolution index. (C) 1998 British Occupational Hygiene Society. Pub lished by Elsevier Science Ltd.