CT-Guided Liver Biopsy: Evaluation of Spectral Data From Dual-Layer Detector CT for Improved Lesion Detection

Evaluation of dual-layer spectral computed tomography (CT) for contrast enhancement during image-guided biopsy of liver lesions using virtual monoenergetic images (VMI) and virtual non-contrast (VNC) images.

Spectral CT data of 20 patients receiving CT-guided needle biopsy of focal liver lesions were used to generate VMI at energy levels from 40 to 200 keV and VNC images. Images were analyzed objectively regarding contrast-to-noise ratio between lesion center (CNR_cent) or periphery (CNR_peri) and normal liver parenchyma. Lesion visibility and image quality were evaluated on a 4-point Likert scale by two radiologists.

Using VMI/VNC images, readers reported an increased visibility of the lesion compared to the conventional CT images in 18/20 cases. In 75% of cases, the highest visibility was derived by VMI-40. Showing all reconstructions simultaneously, VMI-40 offered the highest visibility in 75% of cases, followed by VNC in 12.5% of cases. Either CNR_cent (17/20) or/and CNR_peri (17/20) was higher (CNR increase > 50%) in 19/20 cases for VMI-40 or VNC images compared to conventional CT images. VMI-40 showed the highest CNR_cent in 14 cases and the highest CNR_peri in 12 cases. High image quality was present for all reconstructions with a minimum median of 3.5 for VMI-40 and VMI-50.

When implemented in the CT scanner software, automated contrast enhancement of liver lesions during image-guided biopsy may facilitate the procedure.