Deep Learning-Based Contrast Boosting in Low-Contrast Media Pre-TAVR CT Imaging.

Purpose: This study investigates the impact of deep learning-based contrast boosting (DL-CB) on image quality and measurement reliability in low-contrast media (low-CM) CT for pre-transcatheter aortic valve replacement (TAVR) assessment. Methods: This retrospective study included TAVR candidates with renal dysfunction who underwent low-CM (30-mL: 15-mL bolus of contrast followed by 50-mL of 30% iomeprol solution) pre-TAVR CT between April and December 2023, along with matched standard-CM controls (n = 68). Low-CM images were reconstructed as conventional, 50-keV, and DL-CB images. Qualitative and quantitative image quality were compared among image sets. The aortic annulus was measured by 2 independent readers on low-CM CT images, and interobserver reliability was assessed. Results: DL-CB significantly improved contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) compared to conventional and 50-keV images (CNR: 12.5-13.4, 18-19.8, and 21.9-24; SNR: 10.8-15.5, 10.7-15.5, and 16.8-26.7 on conventional, 50-keV, and DL-CB images, respectively; P < .001). DL-CB achieved comparable CNR (21.9-24 vs 27-27.7, P = .39-.61) and comparable to slightly higher SNR (16.8-26.7 vs 15.7-20.2, P = .003-.80) to standard-CM images. For aortic annular measurement, DL-CB demonstrated high interobserver reliability, with an intraclass correlation coefficient (ICC) of .96 and small mean differences (area: 0.01 cm², limits of agreement [LoA]: -0.52 to 0.55 cm²; perimeter: 0.02 mm, LoA: -4.49 to 4.53 mm). Conclusions: DL-CB improves image quality and provides high measurement reliability in low-CM CT for pre-TAVR assessment in patients with renal dysfunction, without requiring dual-energy CT.