Effect of administered radioactive dose level on image quality of brain perfusion imaging with 99mTc-HMPAO

Abstract

Aim: Brain perfusion imaging by means of 99m Tc-labeled hexamethyl propylene amine oxime (HMPAO) is a wellestablished Nuclear Medicine diagnostic procedure. The administered dose range recommended by the supplying company and reported in bibliography is rather wide (approximately 9.5-27 mCi). This fact necessitates further quantitative analysis of the technique, so as to minimise patient absorbed dose without compromising the examination diagnostic value. In this study, a quantitative evaluation of the radiopharmaceutical performance for different values of administered dose (10, 15, 20 mCi) was carried out. Subsequently, a generic image quality index was correlated with the administered dose, to produce an overall performance indicator. Through this cost-to-benefit type analysis, the necessity of administration of higher radioactive dose levels in order to perform the specific diagnostic procedure was examined. Materials & methods: The study was based on a sample of 78 patients (56 administered with 10 mCi, 10 with 15 mCi and 12 with 20 mCi). Some patients were classified as normal, while others presented various forms of pathology. Evaluation of image quality was based on contrast, noise and contrast-to-noise ratio indicators, denoted CI, NI and CNR respectively. Calculation of all indicators was based on wavelet transform. An overall performance indicator (denoted PI), produced by the ratio of CNR by administered dose, was also calculated. Results: Calculation of skewness parameter revealed the normality of CI, NI and non-normality of CNR, PI populations. Application of appropriate statistical tests (analysis of variance for normal and Kruskal-Wallis test for non-normal populations) showed that there is a statistically significant difference in CI (p 0.05) values. Application of Tukey test for normal populations CI, NI led to the conclusion that CI(10 mCi) = CI(20 mCi) NI(20 mCi), while NI(15 mCi) can not be characterised. Finally, application of non-parametric multiple comparisons showed that CNR(20 mCi)>CNR(10 mCi), while CNR(15 mCi) can not be characterised. Conclusion: Brain perfusion imaging by means of 99m Tc- HMPAO utilising an administered dose of 20 mCi results in improved image quality, on the basis of the estimated indicators and for the range of radioactive dose levels examined. Additionally, this image quality improvement is sufficient to justify the increased radiation burden for the patient.