Congenital heart defects constitute the most common human birth defect, however understanding of how these disorders originate is limited by our ability to model the human heart accurately in vitro. Here we report a method to generate developmentally relevant human heart organoids by self-assembly using human pluripotent stem cells. Our procedure is fully defined, efficient, reproducible, and compatible with high-content approaches. Organoids are generated through a three-step Wnt signaling modulation strategy using chemical inhibitors and growth factors. Heart organoids are comparable to age-matched human fetal cardiac tissues at the transcriptomic, structural, and cellular level. They develop sophisticated internal chambers with well-organized multi-lineage cardiac cell types, recapitulate heart field formation and atrioventricular specification, develop a complex vasculature, and exhibit robust functional activity. We also show that our organoid platform can recreate complex metabolic disorders associated with congenital heart defects, as demonstrated by an in vitro model of pregestational diabetes-induced congenital heart defects.

Congenital heart defects (CHD) constitute the most common birth defect in humans, affecting approximately 1% of all live births. Our ability to understand how these disorders originate is hindered by our limited ability to model the complexity of the human heart in vitro. There is a pressing need to develop more faithful organ-like platforms recapitulating complex in vivo phenotypes to study human development and disease in vitro. Here we report a novel method to generate human heart organoids by self-assembly using pluripotent stem cells. Our method is fully defined, highly efficient, scalable, shows high reproducibility and is compatible with screening and high-throughput approaches. Human heart organoids (hHOs) are generated using a two-step canonical Wnt signaling modulation strategy using a combination of chemical inhibitors and growth factors in completely defined culture conditions. hHOs faithfully recapitulate human cardiac development and are similar to age-matched fetal cardiac tissues at the transcriptomic, structural and cellular level. hHOs develop sophisticated internal chambers with well-organized multi-lineage cell-type regional identities reminiscent of the heart fields and the atrial and ventricular chambers, as well as the epicardium, endocardium, and coronary vasculature, and exhibit functional activity. We also show that hHOs can recreate complex metabolic disorders associated with CHD by establishing the first in vitro human model of diabetes during pregnancy (DDP) to study embryonic CHD. morphological and metabolically effects of increased glucose and insulin, showing the capability of modeling the effects of diabetes during pregnancy (DDP). Our heart organoid model constitutes a powerful novel tool for translational studies in human cardiac development and disease.

Optical coherence tomography (OCT) is widely used for biomedical imaging and clinical diagnosis. However, speckle noise is a key factor affecting OCT image quality. Here, we developed a custom generative adversarial network (GAN) to denoise OCT images. A speckle-modulating OCT (SM-OCT) was built to generate low speckle images to be used as the ground truth. In total, 210 000 SM-OCT images were used for training and validating the neural network model, which we call SM-GAN. The performance of the SM-GAN method was further demonstrated using online benchmark retinal images, 3D OCT images acquired from human fingers and OCT videos of a beating fruit fly heart. The denoise performance of the SM-GAN model was compared to traditional OCT denoising methods and other stateof-the-art deep learning based denoise networks. We conclude that the SM-GAN model presented here can effectively reduce speckle noise in OCT images and videos while maintaining spatial and temporal resolutions.

No abstract

SummaryNano-structures of biological systems can produce diverse spectroscopic effects through interactions with broadband light. Although structured coloration at the surface has been extensively studied, natural spectroscopic contrasts in deep tissues are poorly understood, which may carry valuable information for evaluating the anatomy and function of biological systems. Here we investigated the spectroscopic characteristics of an important geometry in deep tissues at the nanometer scale: packed nano-cylinders, in the near-infrared window, numerically predicted and experimentally proved that transversely oriented and regularly arranged nano-cylinders could selectively backscatter light of the long wavelengths. Notably, we found that the spectroscopic contrast of nanoscale fibrous structures was sensitive to the pressure load, possibly owing to the changes in the orientation, the degree of alignment, and the spacing. To explore the underlying physical basis, we further developed an analytical model based on the radial distribution function in terms of their radius, refractive index, and spatial distribution.