(2024) Augustin | Sharp-kernel PCCT improves lower-extremity CTA vs DSA

Daniel Stromer

In 39 patients with 45 lower-extremity run-off studies, PCCT angiography was compared directly with digital subtraction angiography (DSA) for peripheral arterial occlusive disease. The cohort was challenging, with 77.8% of studies representing Fontaine stage IV disease, making severe calcification and small crural vessels particularly relevant.

Reconstruction kernel selection had a major effect on diagnostic performance. Increasing sharpness from Bv36 → Bv48 → Bv56 progressively reduced calcium blooming and improved vessel sharpness and agreement with DSA, while sensitivity for ≥60% stenosis remained essentially unchanged at approximately 81%.

The benefit was strongest in severely calcified crural arteries: specificity improved from 49.3% with Bv36 to 65.9% with Bv56, while stenosis overestimation decreased and diagnostic confidence increased. Thus, for advanced lower-extremity PAOD, the sharper Bv56 kernel provided the best diagnostic trade-off despite increased image noise.

Access publication

Figure 6. Imaging of the crural vessels in a 95-year-old woman with PAOD stage 4 and necrosis of both thighs. Axial PCD-CTA images of the crural vessels at the level of the distal lower thigh reconstructed with different spatial frequencies on the left side. With Bv36, despite low image noise, the lack of vessel sharpness hampers contrast assessment in the anterior tibial artery. The sharper Bv48 and Bv56 kernels reveal significant wall calcifications and an atheromatous plaque. Meanwhile, reduction of calcium blooming allows for detailed luminal assessment. Corresponding DSA image of the distal run-off confirms a short high-grade stenosis of the distal anterior tibial artery (arrow), while the posterior tibial artery is completely occluded.

Protocol

Detail	Value
Scanner	NAEOTOM Alpha
Scan area	Abdominal aorta and lower-extremity run-off
Scan mode	Standard resolution, QuantumPlus
Acquisition mode	PCCT
Tube voltage [kV]	140
Effective mAs [mAs]	CARE Dose4D automatic
IQ level	26
Dose modulation	CARE Dose4D
CTDIvol [mGy]	6.46 ± 0.80
DLP [mGy·cm]	857.77 ± 120.17
Rotation time [s]	0.5
Pitch	0.4
Matrix	512 × 512
Field of view [mm]	150 × 150 per leg
Detector slice collimation	144 × 0.4 mm
Slice thickness [mm]	0.8
Reconstruction increment [mm]	0.8
Reconstruction kernel	Bv36, Bv48, Bv56
Iterative reconstruction	QIR 3
Reconstruction	Axial, each leg reconstructed separately
Contrast agent	Imeron 350
Contrast volume	90 mL
Saline flush	20 mL pre-bolus + 20 mL post-bolus
Injection rate	3.0 mL/s
Bolus triggering	Abdominal aorta, 140-HU threshold
Trigger delay	18 s

Key Results

Parameter	Main finding
Cohort	39 patients / 45 PCCT run-off studies; 77.8% Fontaine IV
Sensitivity for ≥60% stenosis	Bv36 81.6%; Bv48 81.5%; Bv56 81.0%
Specificity	Increased from 71.1% → 76.9% → 79.6% with Bv36 → Bv48 → Bv56
Severely calcified crural vessels	Specificity improved from 49.3% with Bv36 to 65.9% with Bv56, p = 0.024
Agreement with DSA	Bv56 significantly improved stenosis concordance vs Bv36, p = 0.002
Diagnostic confidence	Highest with Bv56, reflecting reduced calcium blooming and improved vessel sharpness

Source:
Augustin AM, Hartung V, Grunz JP, et al. Photon-Counting Detector CT Angiography Versus Digital Subtraction Angiography in Patients with Peripheral Arterial Disease. Academic Radiology. 2024;31(7):2973–2986. doi:10.1016/j.acra.2024.02.008.