ABSTRACT
X-ray free-electron lasers (XFELs) offer unique capabilities for measuring the structure and dynamics of biomolecules, helping us understand the basic building blocks of life. Notably, high-repetition-rate XFELs enable single particle imaging (X-ray SPI) where individual, weakly scattering biomolecules are imaged under near-physiological conditions with the opportunity to access fleeting states that cannot be captured in cryogenic or crystallized conditions. Existing X-ray SPI reconstruction algorithms, which estimate the unknown orientation of a particle in each captured image as well as its shared 3D structure, are inadequate in handling the massive datasets generated by these emerging XFELs. Here, we introduce X-RAI, an online reconstruction framework that estimates the structure of a 3D macromolecule from large X-ray SPI datasets. X-RAI consists of a convolutional encoder, which amortizes pose estimation over large datasets, as well as a physics-based decoder, which employs an implicit neural representation to enable high-quality 3D reconstruction in an end-to-end, self-supervised manner. We demonstrate that X-RAI achieves state-of-the-art performance for small-scale datasets in simulation and challenging experimental settings and demonstrate its unprecedented ability to process large datasets containing millions of diffraction images in an online fashion. These abilities signify a paradigm shift in X-ray SPI towards real-time capture and reconstruction.
ABSTRACT
BACKGROUND: Antenatal diagnosis of congenital heart defects (CHD) can impact outcomes in neonates with severe CHD. Obstetric screening guidelines and the indications for fetal echocardiography (FE) have evolved in an attempt to improve the early prenatal detection of CHD. Analyzing yield for specific indications will help clinicians better stratify at-risk pregnancies. METHODS: Retrospective cohort study of all FE performed between 2000 and 2010 at a single tertiary care academic medical center in New York City. A total of 9878 FE met inclusion criteria for our study. In cases of multiple gestations (MG), each fetus was counted as a separate study. RESULTS: The number of new diagnosis of fetal CHD by FEs increased 200%. There was a statistically significant increase in those referred for suspected CHD, increased nuchal translucency (NT), MG, and suboptimal imaging (P < .001). The indication of "suboptimal imaging" (SO) not only accounted for 5.23% of all referrals from 2000 to 2002, compared to 22.26% of all referrals from 2008 to 2010 (P < .0001), but also had the lowest yield for diagnoses of CHD (P < .02). CONCLUSIONS: Over the past decade, there has been an increase in utilization of FE with a proportional increase in prenatally diagnosed CHD. For indications such as suspected CHD, NT and MG increases in referrals have led to a proportionate increase in fetal diagnosis of CHD. SO as an indication has the lowest yield of fetal diagnosis of CHD. Antenatal detection of CHD may be improved by a change in obstetric imaging protocols to ensure appropriate referrals.