Abstract
Optical Projection Tomography (OPT) is a powerful three-dimensional imaging technique used for the observation of millimeter-scaled biological samples, compatible with bright-field and fluorescence contrast. OPT is affected by spatially variant artifacts caused by the fact that light diffraction is not taken into account by the straight-light propagation models used for reconstruction. These artifacts hinder high-resolution imaging with OPT. In this work we show that, by using a multiview imaging approach, a 3D reconstruction of the bright-field contrast can be obtained without the diffraction artifacts typical of OPT, drastically reducing the amount of acquired data, compared to previously reported approaches. The method, purely based on bright-field contrast of the unstained sample, provides a comprehensive picture of the sample anatomy, as demonstrated in vivo on Arabidopsis thaliana and zebrafish embryos. Furthermore, this bright-field reconstruction can be implemented on practically any multi-view light-sheet fluorescence microscope without complex hardware modifications or calibrations, complementing the fluorescence information with tissue anatomy.
Highlights
Optical Projection Tomography (OPT) is a powerful three-dimensional imaging technique used for the observation of millimeter-scaled biological samples, compatible with bright-field and fluorescence contrast
We present in-vivo data of Arabidopsis thaliana and zebrafish (Danio rerio) in order to demonstrate that the approach allows one to observe the whole anatomy of unstained organisms
We observe for example that the transverse section of a sample, presents such a poor axial resolution (Fig. 1c, left hand side) that different structures are indistinguishable, hindering any 3D analysis of the sample
Summary
Optical Projection Tomography (OPT) is a powerful three-dimensional imaging technique used for the observation of millimeter-scaled biological samples, compatible with bright-field and fluorescence contrast. The sample is scanned at different positions (up to 100) along the optical axis, and the acquired stack of images is merged into a single entirely focused p rojection[29]. We show that the Bright-Field (BF) contrast can be efficiently reconstructed in 3D by adopting a multiview image processing method: we propose to fuse three-dimensional image data sets of the sample acquired at multiple angles into a single reconstruction.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call