Improvement of epidemiological data analysis by unbiased estimates of log-normal dose distribution Online publication date: Mon, 24-Mar-2008
by Joyeeta Mukhopadhyaya, D. Datta, H.S. Kushwaha
International Journal of Low Radiation (IJLR), Vol. 4, No. 4, 2007
Abstract: The effects of radiation on population and risk assessment are studied through epidemiological studies that are heavily dependent on dose distribution. Radiation doses to individual workers follow a log-normal distribution. Computation of the collective dose with the minimum detectable values (Detection Limit or DL) as zero is biased. Hence the expectation value of the 'missed' doses is computed as the ratio of the collective dose of all dosimeters with DL to the product of the total number of dosimeters. So far, in all epidemiological studies, doses below DL are set equal to the DL itself and yield a positive bias in the dose-response relation. This paper presents an enhanced method by removing the bias with the help of the Expectation Maximisation (EM) algorithm. This algorithm, along with unbiased characteristic estimates of log-normal distribution, significantly improves the estimation of confounders by 95%, and improves the dose-response relationship.
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Low Radiation (IJLR):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email subs@inderscience.com
Our Blog 
Follow us on Twitter 
Visit us on Facebook