摘  要：針對CT值均勻性傳統(tǒng)定義只考慮圖像的中心和外圍部分而不能全面反映整個視野（FOV）內(nèi)CT值變化的缺點，提出一種改進的評價方法，以適應CT新技術質(zhì)量保證檢測的需要。采用直徑為20 cm的圓柱形水模，在GE Discovery CT750 HD掃描儀上選擇常用頭部協(xié)議進行軸向掃描。在水模圖像的3，6，9，12點鐘方向上距離中心0，4.0，8.0 cm處取面積為1 cm2的圓形感興趣區(qū)，分別計算CT值均值和標準差。定義其中最大、最小均值之差為CT值均勻性指數(shù)，最大標準差為噪聲指數(shù)；單能圖像序列中CT值均勻性指數(shù)最大值定義為該序列的CT值均勻性指數(shù)，噪聲指數(shù)最大值為該序列噪聲指數(shù)。結果表明：從FOV中心向外CT值并非單調(diào)變化，CT值均勻性評價新方法更加全面、客觀地評價整個FOV的圖像質(zhì)量，并能反映雙能CT的不同探測器陣列對應圖像之間以及不同能級圖像序列之間的質(zhì)量差異。
關鍵詞：計算機體層攝影；雙能；CT值均勻性；質(zhì)量保證
文獻標志碼：A       文章編號：1674-5124（2016）01-0007-05
A modified evaluation method for CT value uniformity
YANG Kecheng
（School of Biomedical Engineering，Third Military Medical University，Chongqing 400038，China）
Abstract: In current quality assurance （QA） testing guidelines and standards， CT value uniformity （CTVU） is defined only based on image centers and peripheries; instead， the variation in CT value can not be fully reflected within the whole field of view. To solve this problem， a modified method was proposed to meet the QA testing needs of new CT scanners. A 20 cm-diameter water equivalent plastic cylindrical phantom filled with pure water was scanned in a step-and-shot way with a routine dual-energy head protocol （GSI-26） on a GE Discovery CT750 HD scanner. Images were analyzed with a customer designed MATLAB program. Nine circular ROIs （1 cm2） were selected at the spots that were 0 cm， 4.0 cm and 8.0 cm away from the centroid at 3， 6， 9， 12 o’clock positions to calculate the CT mean value and standard deviations. The differences between the maximum and minimum mean value were defined as CTVU indexes; the maximal SD was defined as noise figure; the maximal CTVU index and noise index in virtual monochromatic image sequence were defined as CTVU index and noise index respectively. Experimental results show that the modified evaluation method reflects not only the CT value variation within the whole field of view in a comprehensive and objective manner， but also the quality differences between different images scanned by different detector slices and virtual monochromatic image sequences of dual-energy CT.
Keywords: computed tomography（CT）； dual energy； CT value uniformity； quality assurance