A data-driven approach for spatial alignment of PET add-on detectors with sub-millimeter accuracy

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Instrumentation

Background:

• Clinical Positron Emission Tomography (PET) scanners benefit from high-resolution add-on detectors.
• Precise geometric alignment of these auxiliary detectors is crucial for optimal performance.
• Existing alignment methods may require complex phantoms or extensive calibration.

Purpose of the Study:

• To develop and validate a novel, data-driven methodology for precise geometric alignment of high-resolution add-on detectors with clinical PET scanners.
• To achieve accurate alignment using only point source measurements, simplifying the calibration process.
• To enhance the resolution and sensitivity of PET systems with auxiliary detectors.

Main Methods:

• A data-driven approach was used to calculate transformation parameters by minimizing angular discrepancies in lines of response (LORs).
• The method assumes collinearity between the point source, add-on detector crystal, and scanner crystal.
• An effective depth of interaction (DOI) model was incorporated for improved interaction point estimation, validated via Monte Carlo simulations and experimental measurements.

Main Results:

• The method achieved a geometric discrepancy of 0.23 ± 0.10 mm and 0.20 ± 0.07 mm for two add-on detectors.
• Experimental validation showed centroid deviations consistently below 1 mm across the field of view.
• Angular deviation remained consistently less than 1 degree, demonstrating high alignment accuracy.

Conclusions:

• The proposed alignment method achieves sub-millimeter accuracy, crucial for high-resolution PET imaging.
• The approach is accurate, simple, and does not require physical phantoms.
• This technique promises to significantly improve the quality and diagnostic accuracy of PET systems augmented with auxiliary detectors.