HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Shuke, N. Articles by Sato, J. PubMed Articles by Shuke, N. Articles by Sato, J.

Estimation of coefficients to calibrate standard arterial input function for regional cerebral blood flow mapping with I-123-IMP autoradiographic method: Analysis of error factors

¹ Kojinkai Memorial Hospital, Kushiro, Japan; ; ² Radiology, Asahikawa Medical College, Asahikawa, Japan

995

Objectives: To analyze error factors in the method for estimating calibration coefficient (Cf) of standard arterial input function without arterial blood sampling in I-123-IMP (IMP) autoradiographic (ARG) method (2006 SNM abstract, 47; 362P).

Methods: A total of 165 patients (66±14 y old, M/F=89/76: dementia, 56; degenerative disease, 12; psychiatric disease, 5; vascular disease,71; others, 11 ) were studied with IMP dynamic SPECT. Using brain time-activity curve and uncalibrated standard arterial input function, 2-compartment model analysis was performed to obtain a product of Cf and K1(CfK1), and k2. The quotient of CfK1 and k2 gave CfVd (Vd=K1/k2). Cf could be estimated by correlation equation between CfVd and measured Cf. Five potential error factors that might influence the accuracy of this method: disease categories, degree of CBF reduction, deviation of the estimated model parameters (CfK1 and k2) from their population mean values (Z score) and goodness of fit (correlation coefficient) were investigated in relation to the error fraction of the estimated Cf values.

Results: Deviation of the estimated k2 values (k2 < -2Z: k2<0.0060) was associated with significantly large error fractions (P<0.05). The other factors did not show any significant influence on the error fraction. Excluding 4 cases with large error fraction, correlation equation between measured Cf and CfVD was as follows: Cf=0.047 CfVD +0.208 ( R²=0.634, P<0.0001, n=161).

Conclusions: Accuracy of the proposed method was not influenced by disease categories or by degree of CBF reduction. However, when small k2 values (k2 < 0.0060) was obtained, estimated Cf value might have large error.

Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Shuke, N. Articles by Sato, J. PubMed Articles by Shuke, N. Articles by Sato, J.