Taguchi's analysis to optimize descending aortography for patent ductus arteriosus, with clinical verification.

OBJECTIVE: Taguchi's analysis was adopted to optimize the various factors referring to cardiac angiographic examination used to delineating ductus arteriosus in pediatric patients. SUBJECTS AND METHODS: Thirty-six pediatric patients, 9 male and 27 female, mean age of 6.5±4.7yrs, range 0.6 to 16.6 years were included in the study from January 2004 to April 2005. All patients had patent ductus arteriosus (PDA). Taguchi's L9 orthogonal array was used to generate nine different designs of angiographic levels. Four control factors were selected: a) body surface area (BSA), b) projection angle, c) catheter location, and d) the volume of contrast medium. Each factor was set to three different levels. Statistical analysis, signal-to-noise (S/N) ratio, and analysis of variance (ANOVA) were used to estimate the optimum level of each control factor and to analyze the effect of each factor. RESULTS AND CONCLUSION: The optimal combinations that obtained the highest image quality for PDA were: a) at <0.65m(2) of BSA, b) right anterior oblique (RAO) position at 30(o) plus cranial position (Cr) at 15(o) and lateral view, c) catheter location at T2-3 and d) contrast medium volume: 1.0cc/kg. The projection angle was found to be the most significant factor to delineate ductus arteriosus using the ANOVA test and was not influenced by other factors. The setting of RAO at 30(o) plus Cr at 15(o) and the LAT view obtained the optimal image quality for PDA during descending aortography.