A Comparison of Different Attenuation Correction Methods for Calculating Semiquantitative Parameters of TRODAT-1 SPECT

Background: Chang's method has long been used for attenuation correction for brain imaging. Computed tomography-based attenuation correction (CTAC) is robust and presumed to be the reference standard. Both methods have been used in calculating the semiquantitative TRODAT-1 binding ratio. However, little has been done to compare the semiquantitative parameter produced from these two methods. The aim of this study is to compare the semiquantitative TRODAT-1 binding ratio derived from these two attenuation correction methods. Methods: We retrospectively reviewed Tc-99m TRODAT-1 single photon emission computed tomography/computed tomography (SPECT/CT) visits between July 2017 and December 2017. Each subject's image was attenuation-corrected by both CTAC and Chang's method. We obtained striatal-to-occipital Tc-99m TRODAT-1 binding ratio (SOR) for bilateral caudate nuclei and bilateral putamina from both corrected images. We analyzed the difference between the SORs derived from CTAC and Chang's method by Wilcoxon signed-rank test and Bland-Altman plot. Also, we applied a simple linear regression analysis to study the relationship between the SORs derived from these two attenuation correction algorisms. Results: There were 41 eligible subjects included in our study. The medians of SOR from all four striatal regions showed a significant difference between two attenuation correction methods (p ＜ 0.0001). In Bland-Altman analysis, the biases between two correction methods over four striatal regions were around 0.1, with higher value using Chang's method. All four striatal regions showed excellent fit between the two correction methods with R^2s over 0.9. Finally, the overall striatal SOR corrected by the two methods showed good linear relationship: CTAC SOR = 0.877 × (Chang's method SOR) - 0.01, with R^2 of 0.9630. Conclusion: The SOR differs according to different attenuation correction methods used. The excellent linear relationship between the SORs derived from Chang's method and CTAC makes it possible to convert SORs from one to one, and may be applied to daily practice.

關鍵字

attenuation correction ； Chang's method ； computed tomography ； striatal binding ratio ； TRODAT-1

並列摘要

背景：張氏衰減校正法長久被用在腦影像處理，而電腦斷層（computed tomography, CT）衰減校正法是另一個可靠且標準的方式。兩種方法均有被用於處理多巴胺轉運體（TRODAT-1）的單光子斷層造影（single photon emission CT, SPECT）上，然而，這兩種方法所得到的TRODAT-1半定量數值卻鮮少被比較。本研究的目的是比較兩種不同衰減校正法對TRODAT-1半定量數值之影響。方法：我們從2017年7月至12月回顧性研究接受Tc-99m TRODAT-1 SPECT/CT的病患影像。每位病患的影像都用張氏及電腦斷層衰減校正法做校正。我們在兩種校正法得到的影像上計算紋狀體—枕葉的TRODAT-1結合比率（striatal-to-occipital Tc-99m TRODAT-1 binding ratio, SOR），計算的紋狀體部位包含兩側尾核及兩側殼核四個部位。我們利用Wilcoxon signed rank test和Bland-Altman plot來分析這兩種方法所得出的SOR之間的差異。此外，我們也用簡單線性回歸來分析這兩種方法所得出的SOR之間的線性關係。結果：總共有41位病人的影像納入分析。四個紋狀體部位SOR的中位數在兩種衰減校正法之間有顯著的差異（p ＜ 0.0001）。此外，在Bland-Altman的分析可見四個紋狀體區域的SOR用張氏衰減校正法校正後數值會比電腦斷層校正的數值高出0.1左右。四個紋狀體區域SOR在兩種衰減校正法之間有極好的線性關係（R^2都超過0.9）。最後，總和所有紋狀體區域可見兩種衰減校正法的SOR有很好的線性關係：電腦斷層衰減校正的SOR = 0.877 ×（張氏衰減校正法的SOR）- 0.01，R^2是0.9630。結論：Tc-99m TRODAT-1的SOR數值會因套用不同的衰減校正法而有差異。我們的研究發現兩種衰減校正法之間所計算出的SOR有極好的簡單線性關係，也推導出轉換公式來轉換這兩種衰減校正法所得到的SOR，並用以做為臨床應用。

並列關鍵字

衰減校正；張氏衰減校正法；電腦斷層；紋狀體攝取比率； TRODAT-1

國際替代計量

A Comparison of Different Attenuation Correction Methods for Calculating Semiquantitative Parameters of TRODAT-1 SPECT

全文下載

主題瀏覽