Correlative single-cell hard X-ray tomography and X-ray fluorescence imaging

Abstract

X-ray tomography and x-ray fluorescence imaging are two non-invasive imaging techniques to study cellular structures and chemical element distributions, respectively. However, correlative X-ray tomography and fluorescence imaging for the same cell has yet to be routinely realized due to challenges in sample preparation and X-ray radiation damage. Here we report an integrated experimental and computational workflow for achieving correlative multi-modality X-ray imaging of a single cell. The method consists of the preparation of radiation-resistant single-cell samples using live-cell imaging-assisted chemical fixation and freeze-drying procedures, targeting and labeling cells for correlative x-ray tomography and x-ray fluorescence measurement, and computational reconstruction of the correlative and multi-modality images. With X-ray tomography, cellular structures including the overall structure and intracellular organelles are visualized, while X-ray fluorescence imaging reveals the distribution of multiple chemical elements within the same cell. Our correlative method demonstrates the feasibility and broad applicability of using X-rays to understand cellular structures and the roles of multiple chemical elements and related proteins in signaling and other biological processes.

SwinCell: a transformer-based framework for dense 3D cellular segmentation

“…3D cellular tomography images by holotomography 1,2 , X-ray tomography 2 , confocal microscopy, cryo-electron tomography (cryo-ET) 4,23 , and other emerging imaging techniques are revolutionizing our observation and understanding of cellular structures and processes. Consequently, there is an urgent need for corresponding advances in image analysis algorithms to effectively analyze this data.…”

“…By focusing on smaller, local windows instead of the entire input image, Swin-transformer significantly reduces the complexity of the self-attention mechanism. Traditional transformers exhibit a computational complexity of O(N 2 ) with respect to the input image size N, whereas the Swin-transformer reduces this to O(N) for each window, resulting in lower overall computational demands 24 . In addition, the shift windows mechanism in the Swin-transformer facilitates cellular segmentation in 3D.…”

“…Understanding their behavior, tissue organization, and implications in disease progression and treatment necessitates an exploration of their three- dimensional (3D) spatial context. Various 3D imaging techniques, such as 3D holotomography 1 , X-ray computed tomography 2 , confocal microscopy, cryo-electron tomography (cryo-ET) 3,4 , and volume Electron Microscopy (vEM) 5 have been widely adopted, generating vast amounts of 3D data. Consequently, there is an urgent need for corresponding advances in image analysis algorithms to effectively analyze these 3D data.…”

“…In this work, we developed a 3D SwinCell framework that incorporates the Swin-transformer architecture and 3D U-net to provide cohesive and unified flow predictions, achieving more accurate cell mask prediction and segmentation. The 3D SwinCell algorithm was extensively tested on multimodal datasets, including two publicly available datasets and one in-house dataset collected with Nanolive 3D holotomography 1,2 . The 3D SwinCell algorithm outperforms the state-of-the-art Cellpose algorithm 2,3 in the application of different cell segmentation scenarios.…”

“…The 3D SwinCell algorithm was extensively tested on multimodal datasets, including two publicly available datasets and one in-house dataset collected with Nanolive 3D holotomography 1,2 . The 3D SwinCell algorithm outperforms the state-of-the-art Cellpose algorithm 2,3 in the application of different cell segmentation scenarios.…”