CT值均勻性評(píng)價(jià)方法研究
中國(guó)測(cè)試楊克檉
摘 要：針對(duì)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(gè)視野（FOV）內(nèi)CT值變化的缺點(diǎn)，提出一種改進(jìn)的評(píng)價(jià)方法，以適應(yīng)CT新技術(shù)質(zhì)量保證檢測(cè)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(jìn)行軸向掃描。在水模圖像的3，6，9，12點(diǎn)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(jì)算CT值均值和標(biāo)準(zhǔn)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(biāo)準(zhǔn)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(jié)果表明：從FOV中心向外CT值并非單調(diào)變化，CT值均勻性評(píng)價(jià)新方法更加全面、客觀(guān)地評(píng)價(jià)整個(gè)FOV的圖像質(zhì)量，并能反映雙能CT的不同探測(cè)器陣列對(duì)應(yīng)圖像之間以及不同能級(jí)圖像序列之間的質(zhì)量差異。
關(guān)鍵詞：計(jì)算機(jī)體層攝影；雙能；CT值均勻性；質(zhì)量保證
文獻(xiàn)標(biāo)志碼：A 文章編號(hào)：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
