ISSN:
1741-0444
Keywords:
Deconvolution
;
Renography
;
Retention function
;
Singular value decomposition
;
Total linear least squares
Source:
Springer Online Journal Archives 1860-2000
Topics:
Biology
,
Chemistry and Pharmacology
,
Medicine
Notes:
Abstract By deconvolving the activity/time curves obtained from the blood and kidney, the renal retention function can be calculated, yielding useful clinical data. However, the current deconvolution techniques are not very reliable; they are sensitive to the inaccuracies present in the data. A more reliable, stable and efficient deconvolution technique based on the singular value decomposition, total linear least squares (TLLS), is proposed and its properties are described. The applicability of TLLS as a deconvolution technique in renography is discussed and demonstrated. Results from simulation, as well as from clinical data, are presented to show the advantages of the use of TLLS with respect to noise rejection in the data. They confirm the superiority of TLLS over the current deconvolution techniques. It is concluded that TLLS needs no smoothing and is a powerful, efficient, reliable and stable deconvolution technique.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF02442816