Title :
Small sample maximum likelihood estimation of defibrillation dose-response success curves
Author_Institution :
New Mexico State Univ., Las Cruces, NM, USA
Abstract :
Estimation of the dose-response success rate curve using as few test blocks as possible is essential for automatic implantable defibrillators. The author introduces small sample maximum likelihood estimators of the slope, 50% success point, and the dose-response curve. The logit for the shape of the dose response curve is used as it is convenient for analytical development and has been shown to be representative of canine data. Maximum likelihood estimators of the parameters of a logit function approximation are presented for one and two sample cases. Bias and variance are given for the estimation of the fifty percentile point on the curve for those cases
Keywords :
defibrillators; automatic implantable defibrillators; bias; canine data; defibrillation; dose-response success rate curve; logit function approximation; small sample maximum likelihood estimation; test blocks; variance; Automatic testing; Back; Defibrillation; Electric shock; Heart; Maximum likelihood estimation; Parameter estimation; Probability density function; Rhythm; Surgery;
Conference_Titel :
Computers in Cardiology 1992, Proceedings of
Conference_Location :
Durham, NC
Print_ISBN :
0-8186-3552-5
DOI :
10.1109/CIC.1992.269522