Machine learning derived segmentation of phase velocity encoded cardiovascular magnetic resonance for fully automated aortic flow quantification.

BACKGROUND: Phase contrast (PC) cardiovascular magnetic resonance (CMR) is widely employed for flow quantification, but analysis typically requires time consuming manual segmentation which can require human correction. Advances in machine learning have markedly improved automated processing, but have yet to be applied to PC-CMR. This study tested a novel machine learning model for fully automated analysis of PC-CMR aortic flow. METHODS: A machine learning model was designed to track aortic valve borders based on neural network approaches. The model was trained in a derivation cohort encompassing 150 patients who underwent clinical PC-CMR then compared to manual and commercially-available automated segmentation in a prospective validation cohort. Further validation testing was performed in an external cohort acquired from a different site/CMR vendor. RESULTS: Among 190 coronary artery disease patients prospectively undergoing CMR on commercial scanners (84% 1.5T, 16% 3T), machine learning segmentation was uniformly successful, requiring no human intervention: Segmentation time was < 0.01 min/case (1.2 min for entire dataset); manual segmentation required 3.96 ± 0.36 min/case (12.5 h for entire dataset). Correlations between machine learning and manual segmentation-derived flow approached unity (r = 0.99, p < 0.001). Machine learning yielded smaller absolute differences with manual segmentation than did commercial automation (1.85 ± 1.80 vs. 3.33 ± 3.18 mL, p < 0.01): Nearly all (98%) of cases differed by ≤5 mL between machine learning and manual methods. Among patients without advanced mitral regurgitation, machine learning correlated well (r = 0.63, p < 0.001) and yielded small differences with cine-CMR stroke volume (∆ 1.3 ± 17.7 mL, p = 0.36). Among advanced mitral regurgitation patients, machine learning yielded lower stroke volume than did volumetric cine-CMR (∆ 12.6 ± 20.9 mL, p = 0.005), further supporting validity of this method. Among the external validation cohort (n = 80) acquired using a different CMR vendor, the algorithm yielded equivalently small differences (∆ 1.39 ± 1.77 mL, p = 0.4) and high correlations (r = 0.99, p < 0.001) with manual segmentation, including similar results in 20 patients with bicuspid or stenotic aortic valve pathology (∆ 1.71 ± 2.25 mL, p = 0.25). CONCLUSION: Fully automated machine learning PC-CMR segmentation performs robustly for aortic flow quantification - yielding rapid segmentation, small differences with manual segmentation, and identification of differential forward/left ventricular volumetric stroke volume in context of concomitant mitral regurgitation. Findings support use of machine learning for analysis of large scale CMR datasets.

Screening for Lung Cancer: Communicating With Patients.

OBJECTIVE: The purposes of this article are to detail the experience of a single-center academic institution in applying the patient-centered approach to a lung cancer screening program and to examine how this approach can expand to other aspects of follow-up imaging of lung nodules. CONCLUSION: As the practice of patient-centered radiology gains attention, diagnostic radiologists are findings new ways to become more involved in patient care. A lung cancer screening program is one opportunity for radiologists to consult with and educate patients.

Rare case of pulmonary involvement in an adult with Kawasaki disease.

Kawasaki disease is an acute, self-limited, febrile vasculitis typically seen in early childhood. Pulmonary involvement is uncommon and is not part of the conventional diagnostic criteria. We add to the literature a unique case of a 22year-old male with Kawasaki disease and pulmonary involvement. It illustrates the importance of recognizing unusual presentations of Kawasaki disease and highlights the possibility of pulmonary abnormalities on physical and imaging examination. Awareness of such presentations can help avoid delayed diagnosis, prevent the development of coronary aneurysms, and allow careful observation for imaging resolution.

Pulmonary fibrosis on the lateral chest radiograph: Kerley D lines revisited.

The retrosternal clear space (RCS) is a lucent area on the lateral chest radiograph located directly behind the sternum. The two types of pathology classically addressed in the RCS are anterior mediastinal masses and emphysema. Diseases of the pulmonary interstitium are a third type of pathology that can be seen in the RCS. Retrosternal reticular opacities, known as Kerley D lines, were initially described in the setting of interstitial oedema. Pulmonary fibrosis is another aetiology of Kerley D lines, which may be more easily identified in the RCS than elsewhere on the chest radiograph. TEACHING POINTS: • The RCS is one of three lucent spaces on the lateral chest radiograph. • Reticular opacities in the RCS are known as Kerley D lines. • Pulmonary fibrosis can be seen in the RCS as Kerley D lines. • Kerley D lines should be further evaluated with chest CT.

The King laryngeal tube: a mimic of esophageal intubation in the emergency department.

The King Airway is a temporary airway device used primarily in the pre-hospital setting and typically exchanged for an endotracheal tube upon arrival to the emergency department. Since this usually occurs before imaging, many radiologists are unfamiliar with the King Airway. This lack of familiarity can have important consequences for the patient and treating team. The purpose of this article is to raise awareness of the King Airway among radiologists, emphasize appropriate positioning, and review the imaging complications of incorrect positioning.

One Night in the ED.

Radiology 2.0: One Night in the ED is an app that presents users with scrollable CT images for common scenarios found in any Emergency Department (ED). A great tool for a junior resident before starting call, it allows the user to build confidence by experiencing a simulated environment of viewing all images of a case in real time, rather than being shown screenshots of an abnormality. This can easily be used as one of many resources to prepare residents to spend a night (or more) in the ED.

Scara5 is a ferritin receptor mediating non-transferrin iron delivery.

Developing organs require iron for a myriad of functions, but embryos deleted of the major adult transport proteins, transferrin or its receptor transferrin receptor1 (TfR1(-/-)), still initiate organogenesis, suggesting that non-transferrin pathways are important. To examine these pathways, we developed chimeras composed of fluorescence-tagged TfR1(-/-) cells and untagged wild-type cells. In the kidney, TfR1(-/-) cells populated capsule and stroma, mesenchyme and nephron, but were underrepresented in ureteric bud tips. Consistently, TfR1 provided transferrin to the ureteric bud, but not to the capsule or the stroma. Instead of transferrin, we found that the capsule internalized ferritin. Since the capsule expressed a novel receptor called Scara5, we tested its role in ferritin uptake and found that Scara5 bound serum ferritin and then stimulated its endocytosis from the cell surface with consequent iron delivery. These data implicate cell type-specific mechanisms of iron traffic in organogenesis, which alternatively utilize transferrin or non-transferrin iron delivery pathways.

Endocytic delivery of lipocalin-siderophore-iron complex rescues the kidney from ischemia-reperfusion injury.

Neutrophil gelatinase-associated lipocalin (Ngal), also known as siderocalin, forms a complex with iron-binding siderophores (Ngal:siderophore:Fe). This complex converts renal progenitors into epithelial tubules. In this study, we tested the hypothesis that Ngal:siderophore:Fe protects adult kidney epithelial cells or accelerates their recovery from damage. Using a mouse model of severe renal failure, ischemia-reperfusion injury, we show that a single dose of Ngal (10 microg), introduced during the initial phase of the disease, dramatically protects the kidney and mitigates azotemia. Ngal activity depends on delivery of the protein and its siderophore to the proximal tubule. Iron must also be delivered, since blockade of the siderophore with gallium inhibits the rescue from ischemia. The Ngal:siderophore:Fe complex upregulates heme oxygenase-1, a protective enzyme, preserves proximal tubule N-cadherin, and inhibits cell death. Because mouse urine contains an Ngal-dependent siderophore-like activity, endogenous Ngal might also play a protective role. Indeed, Ngal is highly accumulated in the human kidney cortical tubules and in the blood and urine after nephrotoxic and ischemic injury. We reveal what we believe to be a novel pathway of iron traffic that is activated in human and mouse renal diseases, and it provides a unique method for their treatment.