Dynamic sampling for SAXSTT: towards real-time measurement adaptation.

Area of Science:

• Materials Science
• Nanotechnology
• X-ray Physics

Background:

• Small-angle X-ray scattering tensor tomography (SAXSTT) enables non-destructive 3D nanoscale characterization.
• Current SAXSTT methods require extensive scanning, resulting in lengthy acquisition times.
• Long acquisition times are a significant bottleneck for synchrotron-based experiments.

Purpose of the Study:

• To develop a time-efficient acquisition strategy for SAXSTT.
• To reduce the number of projections needed for accurate 3D nanoscale reconstruction.
• To introduce an adaptive measurement approach compatible with existing systems.

Main Methods:

• Proposed a dynamic sampling strategy with an online error feedback mechanism.
• Implemented adaptive termination of measurements based on real-time reconstruction quality.
• Applied the strategy to experimental SAXSTT datasets.

Main Results:

• Achieved up to a sevenfold reduction in the number of required projections.
• Maintained high reconstruction quality despite reduced data acquisition.
• Demonstrated sample-dependent reduction potential and confirmed strategy robustness via quantitative evaluations.

Conclusions:

• The proposed dynamic sampling strategy significantly enhances SAXSTT efficiency.
• This adaptive approach offers a practical solution for time-limited synchrotron experiments.
• The method is compatible with current beamline infrastructure, facilitating broader adoption.