Interior soft x-ray tomography with sparse global sampling

To investigate the feasibility of interior imaging reconstruction in soft X-ray tomography to achieve higher spatial resolution cellular imaging, including whole-cell imaging.

We develop an alignment and reconstruction algorithm that enables a combination of a low number of images from sparse whole-cell imaging with a high-resolution local interior scan. Based on numerical simulations, we demonstrate that combined reconstructions mitigate the depth of field limitation in high-resolution scans, enable radiation dose optimization, and yield quantitative X-ray absorption values with sparse sampling. Furthermore, we validate our numerical approach using experimental data from two different cell types and demonstrate that combined reconstruction is a reliable method for obtaining high and local spatial resolution within the volume of a whole cell.

The developed sparse reconstruction algorithm provides a robust and faithful visualization of cellular organelles with soft X-ray tomography. A mesoscale imaging approach, such as an interior tomography scan, enables one to "scout" and zoom into the volumes of interest that contain organelles of interest. Utilizing sparse reconstructions, this increase in spatial resolution is achieved without sacrificing larger volume imaging, providing information on the relative position of all organelles within a cell.