A Multi-Parameter Persistence Algorithm for the Automatic Energy Calibration of Scintillating Radiation Sensors

Area of Science:

• Data Analysis
• Signal Processing
• Scientific Instrumentation

Background:

• Traditional peak detection methods are sensitive to parameter choices, hindering reliable interpretation.
• Energy calibration of gamma ray scanners requires robust peak detection.

Purpose of the Study:

• To develop a versatile methodology for robust peak detection in spectral and time-series data.
• To reduce the sensitivity of peak detection to individual parameter selection.

Main Methods:

• Introduced a multi-parameter persistence approach inspired by persistent homology.
• Quantified peak robustness using hyper-volume in parameter space (volumetric multi-parameter persistence - VM-PP).
• Explored parameter space systematically to identify consistently detected peaks.

Main Results:

• VM-PP measure enables robust peak ranking.
• Significantly reduced sensitivity of detection outcomes to parameter selection.
• Reliably differentiated genuine peaks from noise across simulated and experimental data.

Conclusions:

• The VM-PP method enhances interpretability and reliability in complex feature detection.
• Demonstrated effectiveness in spectroscopic calibration and adaptable to diverse signal processing applications.