Association of patient-level characteristics with long-term outcomes after Fontan palliation: Rationale, design, and baseline characteristics of the Pediatric Cardiac Care Consortium Fontan cohort study

BACKGROUND

The Fontan operation is used to palliate single ventricle congenital heart defects (CHD) but poses significant morbidity and mortality risks. We present the design, planned analyses and rationale for a long-term Fontan cohort study aiming to examine the association of patient characteristics at the time of Fontan with post-Fontan morbidity and mortality.

METHODS AND RESULTS

We used the Pediatric Cardiac Care Consortium (PCCC), a US-based, multicenter registry of pediatric cardiac surgeries to identify patients who underwent the Fontan procedure for single ventricle CHD between 1 and 21 years of age. The primary outcomes are in-hospital Fontan failure (death or takedown) and post-discharge mortality through 2022. A total of 1461 (males 62.1%) patients met eligibility criteria and were included in the analytical cohort. The median age at Fontan evaluation was 3.1 years (IQR: 2.4-4.3). While 95 patients experienced in-hospital Fontan failure (78 deaths and 17 Fontan takedown), 1366 (93.5%) survived to discharge with Fontan physiology and formed the long-term analysis cohort. Over a median follow-up of 21.2 years (IQR: 18.4-24.5) 184 post-discharge deaths occurred. Thirty-year post Fontan survival was 75.0% (95% CI: 72.3-77.8%) for all Fontan types with higher rates for current techniques such as lateral tunnel and extracardiac conduit 77.1% (95% CI: 73.5-80.8).

CONCLUSION

The PCCC Fontan study aims to identify predictors for post-Fontan morbidity and mortality, enabling risk- stratification and informing surveillance practices. Additionally, the study may guide therapeutic interventions aiming to optimize hemodynamics and enhance Fontan longevity for individual patients.

Deep-learning-based hepatic fat assessment (DeHFt) on non-contrast chest CT and its association with disease severity in COVID-19 infections: A multi-site retrospective study

Background

Hepatic steatosis (HS) identified on CT may provide an integrated cardiometabolic and COVID-19 risk assessment. This study presents a deep-learning-based hepatic fat assessment (DeHFt) pipeline for (a) more standardised measurements and (b) investigating the association between HS (liver-to-spleen attenuation ratio <1 in CT) and COVID-19 infections severity, wherein severity is defined as requiring invasive mechanical ventilation, extracorporeal membrane oxygenation, death.

Methods

DeHFt comprises two steps. First, a deep-learning-based segmentation model (3D residual-UNet) is trained (N = 80) to segment the liver and spleen. Second, CT attenuation is estimated using slice-based and volumetric-based methods. DeHFt-based mean liver and liver-to-spleen attenuation are compared with an expert's ROI-based measurements. We further obtained the liver-to-spleen attenuation ratio in a large multi-site cohort of patients with COVID-19 infections (D1, N = 805; D2, N = 1917; D3, N = 169) using the DeHFt pipeline and investigated the association between HS and COVID-19 infections severity.

Findings

The DeHFt pipeline achieved a dice coefficient of 0.95, 95% CI [0.93–0.96] on the independent validation cohort (N = 49). The automated slice-based and volumetric-based liver and liver-to-spleen attenuation estimations strongly correlated with expert's measurement. In the COVID-19 cohorts, severe infections had a higher proportion of patients with HS than non-severe infections (pooled OR = 1.50, 95% CI [1.20–1.88], P < .001).

Interpretation

The DeHFt pipeline enabled accurate segmentation of liver and spleen on non-contrast CTs and automated estimation of liver and liver-to-spleen attenuation ratio. In three cohorts of patients with COVID-19 infections (N = 2891), HS was associated with disease severity. Pending validation, DeHFt provides an automated CT-based metabolic risk assessment.

Funding

For a full list of funding bodies, please see the Acknowledgements.

Normalization of breast MRIs using cycle-consistent generative adversarial networks

OBJECTIVES

Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) is widely used to complement ultrasound examinations and x-ray mammography for early detection and diagnosis of breast cancer. However, images generated by various MRI scanners (e.g., GE Healthcare, and Siemens) differ both in intensity and noise distribution, preventing algorithms trained on MRIs from one scanner to generalize to data from other scanners. In this work, we propose a method to solve this problem by normalizing images between various scanners.

METHODS

MRI normalization is challenging because it requires normalizing intensity values and mapping noise distributions between scanners. We utilize a cycle-consistent generative adversarial network to learn a bidirectional mapping and perform normalization between MRIs produced by GE Healthcare and Siemens scanners in an unpaired setting. Initial experiments demonstrate that the traditional CycleGAN architecture struggles to preserve the anatomical structures of the breast during normalization. Thus, we propose two technical innovations in order to preserve both the shape of the breast as well as the tissue structures within the breast. First, we incorporate mutual information loss during training in order to ensure anatomical consistency. Second, we propose a modified discriminator architecture that utilizes a smaller field-of-view to ensure the preservation of finer details in the breast tissue.

RESULTS

Quantitative and qualitative evaluations show that the second innovation consistently preserves the breast shape and tissue structures while also performing the proper intensity normalization and noise distribution mapping.

CONCLUSION

Our results demonstrate that the proposed model can successfully learn a bidirectional mapping and perform normalization between MRIs produced by different vendors, potentially enabling improved diagnosis and detection of breast cancer. All the data used in this study are publicly available at https://wiki.cancerimagingarchive.net/pages/viewpage.action?pageId=70226903.