To evaluate the feasibility of combining Compressed SENSE (CS) with a newly developed deep learning-based algorithm (CS-AI) using a Convolutional Neural Network to accelerate balanced steady-state free precession (bSSFP)-sequences for cardiac magnetic resonance imaging (MRI). 30 healthy volunteers were examined prospectively with a 3 T MRI scanner. We acquired CINE bSSFP sequences for short axis (SA, multi-breath-hold) and four-chamber (4CH)-view of the heart. For each sequence, four different CS accelerations and CS-AI reconstructions with three different denoising parameters, CS-AI medium, CS-AI strong, and CS-AI complete, were used. Cardiac left ventricular (LV) function (i.e., ejection fraction, end-diastolic volume, end-systolic volume, and LV mass) was analyzed using the SA sequences in every CS factor and each AI level. Two readers, blinded to the acceleration and denoising levels, evaluated all sequences regarding image quality and artifacts using a 5-point Likert scale. Friedman and Dunn's multiple comparison tests were used for qualitative evaluation, ANOVA and Tukey Kramer test for quantitative metrics. Scan time could be decreased up to 57 % for the SA-Sequences and up to 56 % for the 4CH-Sequences compared to the clinically established sequences consisting of SA-CS3 and 4CH-CS2,5 (SA-CS3: 112 s vs. SA-CS6: 48 s; 4CH-CS2,5: 9 s vs. 4CH-CS5: 4 s, p < 0.001). LV-functional analysis was not compromised by using accelerated MRI sequences combined with CS-AI reconstructions (all p > 0.05). The image quality loss and artifact increase accompanying increasing acceleration levels could be entirely compensated by CS-AI post-processing, with the best results for image quality using the combination of the highest CS factor with strong AI (SA-CINE: Coef.:1.31, 95 %CI:1.05-1.58; 4CH-CINE: Coef.:1.18, 95 %CI:1.05-1.58; both p < 0.001), and with complete AI regarding the artifact score (SA-CINE: Coef.:1.33, 95 %CI:1.06-1.60; 4CH-CINE: Coef.:1.31, 95 %CI:0.86-1.77; both p < 0.001). Combining CS sequences with AI-based image reconstruction for denoising significantly decreases scan time in cardiac imaging while upholding LV functional analysis accuracy and delivering stable outcomes for image quality and artifact reduction. This integration presents a promising advancement in cardiac MRI, promising improved efficiency without compromising diagnostic quality.

Background/Objectives: To evaluate whether peripheral pulmonary artery and vein morphology improves image-based diagnosis of pulmonary hypertension (PH), in accordance with the recently updated hemodynamic definition. Methods: 229 patients underwent CT pulmonary angiography (CTPA) within 30 days of RHC. Pulmonary vessels ranging between 2 and 10 mm in diameter were extracted and labeled as either arteries or veins by an independently validated fully automated algorithm. Segmentation labels were validated by a radiologist. Results: The segmentation algorithm reached a median accuracy of 90%, aligning with the radiologist’s assessments. Vessel density of pulmonary arteries with diameters between 6 and 10 mm was higher in patients with versus those without PH (median [inter-quartile range]: 8.9 [6.1–10.8] 1/L vs. 6.2 [3.1–7.0] 1/L; p = 0.007). Artery-to-vein ratio was higher in PH (1.32 [0.93–2.06] vs. 0.88 [0.48–1.17], p = 0.004). The artery-to-vein ratio for vessels with diameters between 6 and 10 mm identified PH with an AUC of 0.73 (95% CI: 0.60–0.87). Combining this readout with the DMPA resulted in a numerically higher AUC (sole DMPA AUC: 0.79 (95% CI: 0.68–0.90)) vs. DMPA + artery-to-vein ratio for vessels with diameters of 6–10 mm: 0.81 (95% CI: 0.71–0.92); however, this improvement was not statistically significant (p = 0.4). Conclusions: PH is associated with an increased ratio of peripheral pulmonary arteries to veins within the 6–10 mm diameter range. Pulmonary vascular morphology may complement the established morphological criterion of MPA diameter and improve the diagnostic accuracy of PH on CT.

Background/Objectives: Dual-energy CT (DECT) enables iodine quantification as a snapshot perfusion indicator. Understanding pulmonary iodine distribution in lung-healthy individuals is crucial for clinical applications. This study aimed to automate iodine quantification and assess demographic effects in a lung-healthy reference cohort. Methods: This retrospective cohort study included 112 adults (53% female, mean age 60.3 ± 16.6 years) who underwent repeated portal venous phase chest DECT on a spectral detector dual-layer scanner between 2016 and 2019 at an academic medical center. Patients had dermato-oncological diseases but no visible thoracic tumors. Automatic lung volumetry was merged with reconstructed iodine maps to assess volume and mean iodine concentrations of each lung lobe. Pulmonary iodine perfusion ratios (PIPRs) were calculated by normalizing the pulmonary iodine density against iodine concentration in the portal vein and the main pulmonary artery (mPA). Results: Mean lung volume (f: 3.9 L vs. m: 5.2 L) and iodine concentration (f: 0.87 mg/mL vs. m: 0.69 mg/mL) differed between ages. However, no difference was observed when comparing PIPRs after normalizing against the iodine level in the mPA. PIPR_mPA were consistent across two timepoints (r = 0.88) and decreased with increasing age (≤50 years: 0.18 vs. ≥70 years: 0.15). Conclusions: This study demonstrates that automated pulmonary iodine quantification is feasible. Normalized pulmonary iodine concentration is a more reliable and effective method for evaluating iodine distribution. Our study also highlights the need to account for sex and age variations in future research and clinical applications.