**Introduction**
Reliable visualization of the coronary origins is essential in neonates with Transposition of the Great Arteries (TGA), as coronary anatomy directly influences surgical planning for the arterial switch operation. Photon-counting computed tomography (PCCT) offers high spatial resolution, spectral imaging, and potential dose efficiency, making it promising for neonatal coronary assessment. However, quantitative evidence guiding the choice of electrocardiogram (ECG)-synchronized acquisition strategy, virtual monoenergetic imaging (VMI) level, and reconstruction phase remains limited. This study evaluated which PCCT acquisition and reconstruction strategy provides the most favorable balance between coronary origin visibility, motion robustness, and radiation dose under neonatal cardiac motion.

**Methods**
A neonatal coronary artery phantom was developed to assess small-vessel visibility under controlled cardiac motion. The phantom contained contrast-filled tubes of 1.1 and 2.0 mm diameter, oriented along three spatial directions in a polymethyl methacrylate background. Cardiac motion was generated using an ECG-synchronized motion simulator at 120, 130, 140, and 150 bpm. Two ECG-synchronized PCCT strategies were evaluated: high-pitch flash acquisition and prospective sequential acquisition.

Protocols were compared under matched conditions at 53 keV. Flash acquisitions were additionally reconstructed at 40 and 47 keV to assess VMI effects, while prospective acquisitions were reconstructed at 53 keV across cardiac phases from 25% to 80% of the R–R interval to assess phase influence. Image quality was quantified using segmentation-derived metrics describing tube geometry, cross-sectional area, shape, edge definition, attenuation, spatial overlap, and contrast-to-noise ratio (CNR). Radiation exposure was assessed using scanner-reported CT dose index volume, dose-length product, and size-specific dose estimate.

**Results**
Under matched dynamic conditions at 53 keV, flash and prospective acquisition showed broadly comparable preservation of tube size, cross-sectional area, shape, edge definition, attenuation, and spatial overlap. The clearest protocol-related difference was observed for CNR, which was more favorable for flash acquisition. When the most favorable prospective reconstruction phase was selected, prospective acquisition improved overlap-based preservation.

Within the flash protocol, 40 keV provided the strongest iodine contrast and highest CNR, whereas 53 keV reduced equivalent diameter and cross-sectional area errors. Within the prospective protocol, reconstruction phase was the dominant optimization parameter, with later phases around 70–80% of the R–R interval generally outperforming mid-cycle phases. Prospective acquisition resulted in substantially higher scanner-reported dose indices than flash.

Heart-rate effects were non-monotonic and depended on acquisition timing, reconstruction phase, tube orientation, and z-position. The 1.1 mm tubes represented the most demanding condition, showing greater proportional errors, reduced overlap, lower CNR, and higher segmentation instability.

**Conclusion**
High-pitch ECG-triggered flash PCCT is supported as the preferred practical acquisition protocol for neonatal coronary origin visualization in neonates with TGA. For the primary visual task, 40 keV VMI is preferred because it provides the strongest iodine contrast and highest CNR, while 53 keV remains valuable as a complementary reconstruction when geometric preservation or size-based assessment is important. Prospective acquisition may be reserved for selected cases in which phase-resolved information is expected to improve interpretation sufficiently to justify the additional dose and complexity.
... Read more