(2022) Lennartz | Virtual unenhanced images show limitations for adrenal adenoma characterization

Daniel Stromer

PCCT-derived virtual unenhanced (VUE) images showed no significant difference in mean attenuation compared with true unenhanced (TUE) images for adrenal adenomas (9.2 ± 8.6 HU vs 7.3 ± 8.4 HU; P = .07).

Despite similar mean values, individual VUE measurements showed both over- and underestimation, resulting in 10 of 36 adrenal adenomas being misclassified when the conventional 10-HU threshold was applied to VUE images.

Material decomposition was tissue dependent: VUE underestimated attenuation in muscle and the portal vein while overestimating fat attenuation, indicating that further technical optimization is required before VUE can reliably replace TUE imaging for adrenal adenoma assessment.

Access publication

Figure 2. Case examples for photon-counting detector CT–derived images in adrenal adenoma. Left column: Axial true unenhanced (TUE) images. Middle column: Virtual unenhanced (VUE) images. Right column: Corresponding contrast-enhanced portal venous phase images. (A–C) Images in a 79-year-old woman with lipid-rich adenoma of the left adrenal gland. VUE images led to overestimation of adrenal gland attenuation, which hampered correct classification as lipid-rich adenoma using the 10-HU threshold. (D–F) Images in a 65-year-old woman with lipid-rich adenoma in the left adrenal gland. Classification based on VUE images was in concordance with classification based on TUE images, albeit with a slight overestimation of attenuation on VUE images. (G–I) Images in a 79-year-old woman with lipid-poor adenoma in the left adrenal gland show high contrast material enhancement. VUE images led to 12-HU underestimation when compared with TUE attenuation, which would still result in the correct classification.

Protocol

Detail	Value
Scanner	NAEOTOM Alpha
Scan area	Abdomen / adrenal glands
Scan mode	True unenhanced + portal venous phase
Acquisition mode	QunatumPlus
Tube voltage [kV]	120
Effective mAs [mAs]	Automated exposure control
IQ level	170
Pitch	0.8
Rotation time [s]	0.25–0.5
Detector slice collimation	144 × 0.4 mm
Slice thickness	2 mm
Matrix	512 × 512
Iterative reconstruction	QIR 4
Reconstruction	Axial
Material decomposition	VUE reconstructed from portal venous phase
Contrast agent	Accupaque 300
Contrast phase	Portal venous

Key Results

Parameter	Main finding
Study population	29 patients, 36 adrenal adenomas
VUE vs TUE attenuation	9.2 ± 8.6 vs 7.3 ± 8.4 HU; P = .07
10-HU classification	10/36 adenomas misclassified using VUE
Lipid-rich adenomas	23/36
Lipid-poor adenomas	13/36
Tissue dependence	VUE underestimated muscle/portal vein and overestimated fat attenuation
Clinical implication	VUE cannot yet reliably substitute for TUE when applying the conventional 10-HU threshold

Source: Lennartz S, Schoenbeck D, Kröger JR, Borggrefe J, Niehoff JH. Photon-counting CT Material Decomposition: Initial Experience in Assessing Adrenal Adenoma. Radiology. 2023;306(1):202–204. doi:10.1148/radiol.220919.

Marcel Dijkshoorn

This is a known issue in the literature. Standard Virtual unenhanced VUE has Water/Air/Iodine in the 3-material decomposition leading to a VUE HU overestimation of the fat component. For adrenals LiverVNC with Fat/softtissue/iodine as 3-materials in the decomposition will provide a more accurate adrenal VUE HU.

When measuring on ''water'' VNC higher cutoff values are suggested 13-29HU though not well established yet.

Additionally, with LiverVNC the Fat fraction can be added to provide more information than just VNU HU.

Here 2 publications but there are more on the topic:

https://doi.org/10.2214/AJR.22.28323

Diagn Interv Radiol-28-208-En.pdf