Auto-thresholding for unbiased electron counting

Area of Science:

• Materials Science
• Electron Microscopy
• Detector Physics

Background:

• Advancements in scanning transmission electron microscopy (STEM) necessitate faster and more precise electron detection.
• Electron counting techniques are crucial for structural and functional material characterization.
• Existing retrofittable signal digitizers enhance detector performance but rely on manual thresholding, introducing operator bias.

Purpose of the Study:

• To develop an automated thresholding method for electron counting in STEM.
• To eliminate human bias and improve the efficiency of electron detection.
• To enhance the performance of existing monolithic or segmented electron detectors.

Main Methods:

• An auto-thresholding algorithm was developed to optimize signal digitizer thresholds.
• The method identifies the optimal threshold by maximizing the distinction between electron event signals and noise.
• This approach is applied to retrofittable signal digitizers for electron detectors.

Main Results:

• The auto-thresholding approach successfully determines optimal thresholds for electron counting.
• Implementation leads to easier operation and increased throughput for electron detection.
• Human bias in signal digitisation is effectively eliminated.

Conclusions:

• Automated thresholding provides a more objective and efficient method for electron counting.
• This technique improves the accessibility and performance of fast scanning transmission electron microscopy.
• The developed method supports advanced materials characterization at high speeds.