Cardiac CT has become a pivotal tool in cardiovascular diagnostics. However, imaging the constantly moving heart remains technically challenging due to intrinsic cardiac motion, patient-specific variability, and hardware limitations—all of which can introduce artifacts. Accurate phase selection is critical to address these issues during image reconstruction. CardioXphase, an AI-powered algorithm, automatically identifies the optimal phase for each cardiac cycle, minimizing motion artifacts and producing sharper coronary images for more reliable diagnoses. By automating this previously manual and time-consuming task, it simplifies workflow and reduces reliance on subjective radiologist input.

Complexity of Coronary Motion Brings Challenges to Conventional Reconstruction

Coronary artery motion is influenced by more than just motion amplitude—it also depends on heart rate variability and vessel anatomy, all of which contribute to complex and dynamic shifts in the quiescent phase of each vessel within the cardiac cycle. The right coronary artery, for instance, often exhibits residual motion due to its long, curved course, even during globally quiescent phases. In patients with elevated or irregular heart rates, the optimal reconstruction phase can vary unpredictably from beat to beat. These complexities highlight the need for vessel- and cycle-specific phase selection, offering more consistent and diagnostically reliable coronary imaging than a one-size-fits-all approach. [1][2].

CardioXphase

CardioXphase is a detector coverage independent algorithm that leverages advanced AI to deliver the best possible coronary image quality. By focusing on vessel clarity analysis, it provides consistently sharp and reliable coronary visualization tailored to clinical needs.

Intelligent Evaluation for Vessel Quality

Conventional methods select phases with minimal cardiac motion, but these do not always yield the best image quality for coronary arteries. By reconstructing multiple phases rapidly, CardioXphase automatically extract the vessels using AI, and evaluates each phase based on coronary sharpness and circularity, selecting the one with the highest quality score. This AI-driven approach achieves image quality on par with experienced radiologists by focusing specifically on coronary features.

Dynamically Adapt to Each Cycle

For CT systems with the detector coverage of less than 16cm, data acquisition of the whole heart requires multiple cardiac cycles. This implies that each cardiac cycle will have a corresponding optimal phase, especially in the case of irregular heart rhythms. CardioXphase chooses the optimal phase independently from each cardiac cycle, improving the overall coronary artery image quality.

Interactive Options for Imaging Intent

CardioXphase AI algorithm enables automatic identification of the optimal phase across the full cardiac cycle—or focus specifically on systole or diastole. It also enables targeted phase selection for the left and right coronary arteries individually. Designed for those who need more tailored views and detailed evaluations.

Scan:
CorCTA Helical
0.25 sec/rotation
kV: 100
CTDIvol: 13.91 mGy
Eff.Dose: 4.49 mSv
Avg HR: 59 bpm
Min HR: 56 bpm
Max HR: 63 bpm

Contrast:
350 mgl/ml
60 ml
4.5 ml/s

Recon:
Matrix: 512 * 512
Slice: 0.5 * 0.5mm
HIR: C_SOFT_AA/KARL 9

Scan Parameters:
Coronary CTA
kV: 100 mAs: 166
0.25 s/rotation
CTDIvol: 13.58 mGy
ED: 2.66 mSv

AvgHR: 82 bpm
MinHR: 78 bpm
MaxHR: 87 bpm

Contrast:
55 ml, 4.0 ml/s

Recon:
C_SOFT_AA, KARL: 8
0.5 mm * 0.5 mm
WW/WL: 800/200

Scan:
CorCTA Helical
0.25 sec/rotation
Coronary CTA
kV: 100
mAs: 140
CTDIvol: 20.215 mGy
Eff.Dose: 4.21 mSv

AvgHR: 61 bpm
MinHR: 59 bpm
MaxHR: 82 bpm

Contrast:
350 mgl/ml
50 ml
5.0 ml/s

Recon:
Matrix: 512 * 512
Slice: 0.75 * 0.5mm
Kernel: C_SOFT_BA/KARL 5

The ePhase (CardioXphase) algorithm offers a marked improvement over traditional manual or default phase selection methods in coronary CT angiography (CCTA). It consistently delivers diagnostic-quality images by accurately identifying optimal reconstruction phases without the need for radiologist intervention. Unlike manual selection, which often relies on trial-and-error using phase previews on a single slice, ePhase (CardioXphase) evaluates the entire coronary vessel to determine the best phase, ensuring greater objectivity and reducing variability.

—— < Image quality of automatic coronary CT angiography reconstruction for patients with HR ≥ 75 bpm using an AI-assisted 16-cm z-coverage CT scanner. >
BMC Med Imaging 21, 24 (2021)

1. van der Wall EE, de Graaf FR, van Velzen JE, Jukema JW, Bax JJ, Schuijf JD. 320-row CT: does beat-to-beat motion of the coronary arteries affect image quality? Int J Cardiovasc Imaging. 2012 Jan;28(1):147-51. doi: 10.1007/s10554-010-9794-x. Epub 2011 Jan 30. PMID: 21279691; PMCID: PMC3275735.
2. Automatic Determination of Differential Coronary Artery Motion Minima for Cardiac Computed Tomography Optimal Phase Selection Lessick, Jonathan et al. Academic Radiology, Volume 22, Issue 6, 697 - 703