PhD Scientific Days 2024

Budapest, 9-10 July 2024

Poster Session T - Cardiovascular Medicine and Research 3.

Optimization of ultrahigh-resolution coronary CT angiographic images’ reconstructions using ex vivo model


Lili Száraz1, Dalma Tóth2, Ibolyka Dudás, MD1, Sámuel Beke, MD1, Emese Zsarnóczay, MD1, Attila Fintha, MD, PhD3, Gergely Rácz, MD, PhD3, Pál Maurovich-Horvat, MD, PhD, MPH, DSc1
1: Medical Imaging Centre
2: Semmelweis University, Faculty of Medicine
3: Department of Pathology and Experimental Cancer Research

Text of the abstract

Introduction: The latest photon-counting detector computed tomography (PCD-CT) technology offers the option of ultra-high-resolution (UHR) scan mode with increased diagnostic accuracy. However, the extent of this effect and post-processing optimizing protocols in the clinical coronary CT angiography (CCTA) routine remains unclear.
Aims: We aimed to investigate the feasibility of the helical and sequential CCTA scan modes directly with PCD-CT in an ex vivo study. We aimed to optimize image reconstruction in case of a high coronary calcium load.
Methods: Imaging in helical and sequential UHR scan modes were repeated on ex vivo human hearts. Eight different vascular kernels were used for reconstruction (Bv40 up to Bv89) with matrix sizes of 512X512 and 1024X1024. Image noise (IN), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), vessel sharpness (VS), and blooming artifact were quantified. Subjective image quality (IQ) was assessed.
Results: Five ex vivo hearts were included, and eight calcified plaques were signed for analysis. The IN, SNR, CNR, VS, and blooming artifacts differed significantly (p<0,005). Softer kernels showed the lowest IN, VS and highest SNR and CNR values in both acquisition modes (IN(Bv40)[hel]: 30,6±7,3 HU; SNR(Bv40)[hel]: 12±6,9 HU; CNR(Bv40)[hel]: 16,2±8,4 HU; VS(Bv40)[hel]: 421,7±72,4 ΔHU/mm versus IN(Bv40)[seq]: 26,7±11,8 HU; SNR(Bv40)[seq]: 15,5±3,7 HU; CNR(Bv40)[seq]: 17,1±6,2 HU; VS(Bv40)[seq]: 443,3±22,6 ΔHU/mm). Sharper kernels achieved the opposite results. Comparing quantitative parameters between helical and sequential modes, no significant difference existed for any kernel (all p>0.05). Blooming artifacts reduced towards Bv64[seq] and Bv72[hel] (35,2±10,2% versus 33,±8,4%). Subjective IQ, IN, and VS were rated excellent for kernels Bv40[seq/hel] up to Bv64[seq/hel].
Conclusions: UHR helical and sequential imaging modes of a PCD-CT do not differ significantly in IQ and quantitative parameters despite elevated radiation dose in helical scan mode. Optimal IQ and vessel delineation were achieved using vascular kernel Bv64 with matrix size 512X512 in both modes.
Funding: „Project no. 2023-2.1.2-KDP-2023-00016 has been implemented with the support provided by the Ministry of Culture and Innovation of Hungary from the National Research, Development and Innovation Fund, financed under the KDP-2023 funding scheme.”