Advantages Of Gemstone Spectral Imaging (GSI) In The Diagnosis Of Hypovascular Liver Lesions

Objective To evaluate the value of Gemstone spectral imaging (GSI) in the diagnosis of hypovascular liver cancer. Methods Gemstone spectral imaging (GSI) CT data of34cases with hypovascular liver neoplasms were retrospectively studied. GSI Viewer was applied to acquire iodine-based image,65keV monochromatic images and conventional polychromatic images (140kVp). The number of lesions, contrast ratio (CR), singal-to-noise ratio (SNR),contrast-to-noise ratio (CNR) and subjective image quality were calculated in the same slice, Paired t tests were used to compare differences between the two modes. Results Only12lesions could be found in conventional polychromatic images while monochromatic images revealed34lesions. compared with polychromatic images,the CR, CNR and subjective image quality were higher in monochromatic images.(21.19±15.23) vs.(11.13±10.09), t=7.05,P<0.01;(1.53±0.88)vs.(0.88±0.68),t=7.19,P<0.01;(4.32±0.65)vs.(3.7±0.53),t=4.49,P<0.01. SNR was no statistical difference between65keV monochromatic images and conventional polychromatic images.(14.02±2.08) vs.(13.64±2.07),t=0.72,P>0.05. Conclusion GSI65keV monochromatic images could significantly increase the detection of hypovascular liver neoplasms. Objective To study which the optimal spectrum of keV to detect hypovascular liver lesion with spectral CT monochromatic imaging. Materials and Methods The82patients’data with hypovascular liver lesion during the arterial phase (AP), portal venous phase (PVP) and hepatic venous phase (VP) in enhanced CT scans were applied with CT spectral imaging technique, all data were sent to GE Advanced Workstation (AW4.4), use the GSI Viewer1.0for data post-processing analysis. Use the GSI mode to reconstruct the monochromatic imaging data (40-90keV), interval of5keV in each group, In the same slice,lesions and liver tissues CT values, image noise and subjective imaging quality scores were measured with a same size of the region of interest(ROI). The value difference between CT lesions and liver tissues were calculated. All export data using SPSS17.0statistical software for statistical analysis, each group data were expressed as mean±standard deviation, and the correlation between image noise and subjective imaging quality score using Pearson’s correlation analysis for statistical analysis. Result The image background noise lower keV levels were located at the level of50keV and70keV in the arterial phase, portal venous phase and hepatic venous phase, at the50keV level monochromatic image background noise in arterial phase, portal venous phase, venous phase were20.06±3.49HU,24.38±7.74HU,23.44±5.49HU, at the70keV level monochromatic image background noise in the three phases were11.58±1.74HU,13.93±4.40HU,13.21±3.07HU; the value difference between lesions and normal liver tissue gradually reduced while the keV increased. At the50keV level monochromatic image arterial phase, portal phase and venous phase contrast were12.41±3.66HU,47.35±115.35HU,31.82±9.54HU;and70keV monochromatic image in the three phases were0.07±0.02HU,31.51±9.02HU,23.30±5.31HU;50keV level monochromatic image Contrast-to-Noise-Ratio (CNR)in arterial phase, portal venous phase and venous phase were0.62±0.04,1.94±0.18,1.36±0.11,70keV level CNR in three phase were0.11±0.01,2.26±0.19,1.76±0.18; Image background noise and subjective imaging quality score showed a significant negative correlation (arterial phase(AP), r=0.19,p<0.05; portal venous phase (PVP),r=0.71, p<0.05; venous phase(VP), r=0.73, p<0.05). Conclusions Reconstruct50keV monochromatic imaging data in the arterial phase could obtain the largest CNR, the optimal monochromatic level was determined at50keV, And the portal venous phase and hepatic vein phase reconstruct70keV level image data lesions with normal liver tissue has maximum contrast and minimum noise. With a maximum contrast to noise ratio(CNR), conducive to the detection and diagnosis of the hypovascular liver lesions, and assist in improving diagnostic confidence.