Biomedical subjects

F N Madden

Publications and source records attributed to F N Madden.

3 recordsLinked to original sources

PubMed1998

CODE: a deconvolution program implementing a regularization method of deconvolution constrained to non-negative values. Description and pilot evaluation.

A regularization method of deconvolution constrained to non-negative values is described. The method gives smooth estimates of the input function whilst providing a feasible fit (in terms of least squares) to measurements. A description of the program CODE (constrained deconvolution) which implements the method is given. A new methodology for a pilot evaluation of deconvolution programs is also proposed. The methodology is based on synthetic data. It employs a variety of shapes of the input function, low (1%) and high (15%) values of the measurement error, and incorporates primary (accuracy) and secondary (bias) performance measures. The performance of CODE is evaluated and it is suggested that CODE provides estimates of the input function with acceptable accuracy.

Data Interpretation, Statistical↗

PubMed1998

Nonlinear pharmacokinetics of tissue-type plasminogen activator in three animal species: a comparison of mathematical models.

A recent study presented plasma concentrations of tissue-type plasminogen activator in three different animal species and at several different dose levels. A three-compartment mammillary model with capacity-limited elimination (of Michaelis-Menten form) was postulated to describe the data. In the present study, several alternative model structures are examined with the view of determining whether better fits can be obtained, whether linear models are significantly worse than nonlinear models, and whether all three compartments are really necessary.

Animals↗

PubMed1996

A comparison of six deconvolution techniques.

We present results for the comparison of six deconvolution techniques. The methods we consider are based on Fourier transforms, system identification, constrained optimization, the use of cubic spline basis functions, maximum entropy, and a genetic algorithm. We compare the performance of these techniques by applying them to simulated noisy data, in order to extract an input function when the unit impulse response is known. The simulated data are generated by convolving the known impulse response with each of five different input functions, and then adding noise of constant coefficient of variation. Each algorithm was tested on 500 data sets, and we define error measures in order to compare the performance of the different methods.

Algorithms↗