Detection of low-contrast lesions in computed body tomography: an experimental study of simulated lesions.

Abstract

Observer accuracy in the identification of low-contrast objects in computed tomography (CT) was studied. Thresholds were established for detection of lesions of various sizes and attenuation differences in CT images produced at different radiation doses. Noise reduction was important in identifying certain types of lesions. Detection was not accurate when the standard deviation of the mean of an organ exceeded the difference in the means of the lesion and the surround region.

Contrast-to-Noise Ratio and Low-Contrast Object Resolution on Full- and Low-Dose MDCT: SAFIRE Versus Filtered Back Projection in a Low-Contrast Object Phantom and in the Liver

“…In a very early investigation of a singledetector CT scanner (Technicare Delta-scan 2020, Ohio-Nuclear; collimation unknown, but likely 10 mm) in 1980, Meaney et al [47] …”

Feasibility of Computed Tomography Pulmonary Angiography with Low Flow Rates

“…A threshold for the vascular enhancement of ≥200 HU was used for good opacification. This threshold was chosen following the calculations used by Wittram, which was based on observations by Meanly et al[36] They proposed a minimum contrast enhancement of pulmonary arterial blood of 93 HU to detect all acute emboli. [3] As contrast enhancement get worse toward the periphery, some researchers claim higher HU values for acute emboli.…”

Computed tomography in the detection of small hepatocellular carcinomas

“…A small lesion is invisible by minimal density difference between the lesion and the normal liver parenchyma, partial volume effect by overlapping of air, and noise of the lesion [12,13]. During the survey of precontrast CT, control of the position of window center and width contribute to detection of small low-density lesions, although these settings may cause false positive interpretation [12,14].…”