Efficient exponential regression with exact fiducial limits to fit cardiac pressure data

Many observations in the biological field can adequately be fitted to an exponential regression equation, commonly in the form y = alpha + beta eyx. However, the exact confidential limits of the parameters remain a problem. An algorithm to solve this problem is proposed here and tested with experimental data obtained from ventricular pressure measurements in isolated working small animal hearts. A subroutine based on this algorithm is written in Borland's Turbo-Pascal but easily portable to other languages.