Measuring kidney depth of Chinese people with kidney dynamic imaging

Wang Rong-fu
2007-01-01
Abstract:Objective To explore a better formula to calculate the kidney depth of the Chinese people in kidney dynamic imaging. Methods One hundred and forty-seven cases were selected from the patients who undertook abdomen CT scanning from December, 2004 to April, 2005, and their sex, age, height and weight were recorded. Kidney depth was determined by measuring the depth from the skin to the anterior and posterior surfaces of the kidney at the renal hilum and then the average of these values were taken as the kidney depth. A multiple linear stepwise regression analysis was carried out to determine the relative importance of each of several variables in order to develop a new regression equation for estimating right and left kidney depth. The 147 cases were divided into two groups randomly, of which the first group was used to derive a regressive formula, and the second was used to verify the formula. Finally, all the cases together were studied to derive a formula to calculate the kidney depth. The comparison between the two methods was conducted by applying Bland-Altman method. Results The depth regression equations deduced from this study were as followed: right kidney depth (cm)=15.449×(weight/height)+0.009637×age+0.782, left kidney depth (cm)=16.772×(weight/height)+0.01025×age+0.224 (weight: kg, height: cm). The equations was more accurate than Tonnesen formula. Conclusion In this research, with the kidney depth of CT as reference value, the new formula was obtained for the Chinese people in kidney dynamic imaging, and it was more accurate than Tonnesen formula.
What problem does this paper attempt to address?