In-Stent Restenosis in Peripheral Arterial Disease: Ultra-High-Resolution Photon-Counting Versus Third-Generation Dual-Source Energy-Integrating Detector CT Phantom Study in Seven Different Stent Types

Abstract

Purpose: The visualization of peripheral in-stent restenosis using energy-integrating detector CT is challenging due to deficient spatial resolution and artifact formation. This study compares the first clinically available photon-counting detector CT to third-generation dual-source energy-integrating detector CT.

Materials and methods: Nylon cylinders with central bores (4 mm, 2 mm), mimicking 75% and 95% stenoses, were placed inside seven different 8-mm diameter stents and filled with diluted contrast medium. Phantoms were scanned with photon-counting detector CT at slice thicknesses of 0.2 mm (available only in this scanner type), 0.5 mm, and 1.0 mm versus 0.5 mm and 1.0 mm in energy-integrating detector CT at matched CT dose indices. Contrast-to-noise ratios were calculated from attenuation rates. Residual lumen size was measured as full width at half-maximum. Subjective image quality was assessed by two independent blinded raters.

Results: Mean contrast-to-noise ratio was lowest in photon-counting detector CT at 0.2 mm slice thickness (0%, 75%, and 95% in-stent restenosis: 6.11 ± 0.6, 5.27 ± 0.54, and 5.02 ± 0.66) and highest at 1.0 mm slice thicknesses with similar measurements in photon-counting detector CT and energy-integrating detector CT (11.46 ± 1.08, 9.94 ± 1.01, 8.26 ± 1.0 vs. 3.34 ± 1.0, 9.92 ± 0.38, 7.94 ± 1.07). Mean full width at half-maximum measurements in photon-counting detector CT at 0.2 mm slice thickness for 0%, 75%, and 95% in-stent restenosis were 8.00 ± 0.37, 3.98 ± 0.34, and 1.92 ± 0.16 mm. Full width at half-maximum was least precise in 95% in-stent restenosis at 1.0 mm slice thickness with similar measurements between scanners (1.57 ± 0.33 vs. 1.71 ± 0.15 mm). Interrater correlation coefficient was 0.75 [95% CI: [0.53; 0.86]; subjective scores were best at 0.2 mm slice thickness in photon-counting detector CT (19.43 ± 0.51 and 19.00 ± 0.68).

Conclusion: In phantom in-stent restenosis in 8 mm stents, we observed similar full width at half-maximum for photon-counting detector CT and energy-integrating detector CT in 0% and 75% in-stent restenosis, but at 95% in-stent restenosis, FWHM tended to be more accurate in smaller slice thicknesses in both scanners. Subjective image assessment yielded best results at 0.2 mm slice thickness in photon-counting detector CT despite lower contrast-to-noise ratio.

Keywords: Energy-integrating detector CT; In-stent restenosis; Phantom study; Ultra-high-resolution photon-counting detector CT.

Fig. 1

a Peripheral artery in-stent restenosis phantom. I = Stenosis outside stent; II = stenosis inside stent; III = stent without stenosis. Stents with a diameter of 8 mm (length varying between 40 and 60 mm) were placed inside plastic tubes (outer diameter = 10 mm, inner diameter = 8.5 mm). Cylindrical solid silicone stenosis (length = 30 mm, diameter = 8 mm) with residual lumina of 4 and 2 mm, respectively, was placed on opposite sides of the stent. Tubes were filled with diluted contrast medium, mimicking contrast-enhanced blood and sealed with silicone. Phantoms were then placed inside a water-filled container measuring 36 × 26 × 14 cm, imitating surrounding tissue. b Measurement of full width at half-maximum in a plot profile of a 95% in-stent restenosis. FWHM = full width at half-maximum. The maximum and minimum (average by area) of the plot profile were measured to identify the half-maximum. FWHM is measured between the crossings of the curve at half-maximum. c Image of in-stent restenosis phantom using ultra-high-resolution photon-counting detector computed tomography. ROI = region of interest. Attenuation rates of six specific ROI were measured: inside the non-stenosed stented lumen (ROI 1); inside the 75% stenosis of the stented segment (ROI 2); inside the lumen of the 75% stenosed stented segment (ROI 3); inside the 95% stenosis of the stented segment (ROI 4); inside the lumen of the 95%-stenosed stented segment (ROI 5); and the standard deviation of density outside the phantom indicating image noise (ROI 6)

Fig. 2

Boxplots of contrast-to-noise ratio and full width at half-maximum measurements at 0%, 75% and 95% stenosis EID-CT = energy-integrating detector computed tomography; FWHM = full width at half-maximum; PCD-CT = photon-counting detector computed tomography; CNR = contrast-to-noise ratio. White boxplots represent PCD-CT measurements, whereas gray boxplots represent measurements of EID-CT. The box shows the first quartile (bottom), median (bold line within the box) and the third quartile (top). Whiskers extend to the minimum and maximum. Outliers are defined as observations that are outside the interval [first quartile–1.5 × interquartile range; third quartile + 1.5 × interquartile range]. Outliers are shown as circles and in the case of outliers, the whiskers extend to the smallest/largest observation within that interval. In addition, the individual data points are shown in the graphics. Coincident points are stacked

Fig. 3

Visual impression of in-stent restenosis Comparison of an in-stent restenosis phantom with a stent caliber of 8 mm between (from left to right) energy-integrating detector CT at 1.0 and 0.5 mm slice thickness vs. photon-counting detector CT at 1.0, 0.5, and 0.2 mm slice thickness