We analyzed the baseline data to assess the association between the qualitative measure of GWD (mean GWD ratings provided by 2 readers) and quantitative measures (conspicuity and CNR) using regression analysis. 18 ⇓– 20 Using both qualitative and quantitative measures such as conspicuity and contrast-to-noise ratio (CNR), we aimed to test whether CIE improved gray-white differentiation (GWD) in pediatric head CT images with normal findings. 13Ĭorrelative Image Enhancement (CIE, patent pending) is an image-processing algorithm that aims to increase the conspicuity of details of interest within digital images by targeted noise reduction and contrast optimization. 10, 11, 15 ⇓– 17 In addition, some image postprocessing techniques have been proposed that can improve gray-white differentiation on head CT images after they have been generated. 10 ⇓ ⇓ ⇓ ⇓ ⇓ ⇓– 17 Many of these strategies such as tube current modulation or use of iterative reconstruction techniques are applicable at the level of image generation or reconstruction at the scanner. 4 ⇓ ⇓ ⇓ ⇓– 9 A number of optimization techniques have been developed that aim to maintain diagnostic quality while reducing radiation exposure. This is the case with pediatric head CTs, which, in response to increased awareness and concern about radiation-induced harm, are often acquired using aggressive dose reduction. 3 The ease of gray-white differentiation is adversely affected by the presence of noise, which, while being inherent in the process of image generation, can be particularly prominent when images are acquired at a lower radiation dose. 1 ⇓– 3 Accordingly, it has long been used as a measure of the diagnostic quality of head CT images. ABBREVIATIONS: CIE Correlative Image Enhancement CNR contrast-to-noise ratio GWD gray-white differentiation SI signal intensityĭifferentiation of gray and white matter on head CTs is important in the recognition of normal structures as well as identification of pathologic changes within brain parenchyma.

Osirix lite length tool shows 0 pixels