Local regularization of tilt projections reduces artifacts in electron tomography

Area of Science:

• Cellular Biology
• Microscopy
• Image Processing

Background:

• Electron tomography (ET) provides high-resolution 3D reconstructions of cellular components.
• Image artifacts, particularly from electron-dense regions, degrade tomogram quality and obscure structures.
• Existing methods struggle with artifacts that hide proximal biological details.

Purpose of the Study:

• To develop and evaluate a pre-reconstruction filter for reducing artifacts in ET.
• To improve the quality of 3D tomograms, especially in the presence of electron-dense materials.
• To enhance the visualization of cellular structures in ET data.

Main Methods:

• Introduction of a non-conservative non-linear isotropic diffusion (NID) filter.
• Application of the NID filter to tilt projections before reconstruction.
• Testing on simulated data and real samples of malaria parasite-infected red blood cells.

Main Results:

• The NID filter effectively identifies and reduces local irregularities in tilt projections.
• Significant improvement in the quality of reconstructed tomograms was observed.
• Suppression of artifacts from electron-dense regions, enabling better visualization of underlying structures.

Conclusions:

• The proposed NID filter is a valuable tool for pre-processing ET data.
• This method enhances the structural and functional investigation of cellular components.
• Improved tomogram quality facilitates research in fields like malaria parasite biology.