Real-time reconstruction and visualisation towards dynamic feedback control during time-resolved tomography...

Area of Science:

• Materials Science
• Biophysics
• Synchrotron Science

Background:

• Synchrotron-based X-ray microscopy enables high-resolution 3D structural studies of dynamic systems.
• Rapid data acquisition in time-resolved experiments generates large datasets, overwhelming computational resources for immediate processing.
• Current methods often result in 'blind' data acquisition, hindering real-time monitoring and optimization.

Purpose of the Study:

• To develop an efficient system for real-time reconstruction, visualization, and analysis of 3D X-ray microscopy data.
• To enable on-the-fly assessment of experimental conditions and data quality during dynamic studies.
• To facilitate adaptive feedback control for time-resolved in situ tomographic experiments.

Main Methods:

• Implementation of a real-time data processing pipeline using a single computing workstation.
• Focus on reconstructing and analyzing a small number of arbitrarily oriented slices.
• Integration with the TOMCAT beamline at the Swiss Light Source.

Main Results:

• Achieved reconstruction throughput matching data acquisition rates (multiple slices per second).
• Enabled real-time visualization and analysis of dynamic 3D structural changes.
• Demonstrated the feasibility of monitoring experiments as they happen.

Conclusions:

• The developed system overcomes computational bottlenecks in time-resolved X-ray microscopy.
• Real-time analysis allows for immediate identification of issues and optimization of experimental parameters.
• This advancement is crucial for adaptive feedback control in dynamic in situ experiments.