Towards monolithic scintillator based TOF-PET systems: practical methods for detector calibration and operation.

Area of Science:

• Medical Imaging
• Nuclear Physics
• Materials Science

Background:

• Monolithic scintillator crystals offer excellent spatial resolution (<2mm FWHM) and timing ( <200ps FWHM) for time-of-flight positron emission tomography (TOF-PET).
• Current position and time estimation methods for these detectors are lengthy and complex, hindering clinical TOF-PET applications.
• High sensitivity and depth-of-interaction (DOI) information are crucial for advanced PET imaging.

Purpose of the Study:

• To revise and accelerate algorithms for practical calibration and operation of monolithic scintillator detectors in TOF-PET systems.
• To improve the efficiency of position and time-of-interaction estimation.
• To enhance system sensitivity by incorporating data from events with missing information.

Main Methods:

• Accelerated k-nearest neighbor (k-NN) classification for x,y-position estimation, reducing computation time by ~200x.
• Revised DOI and time-of-interaction estimation procedures enabling calibration with fan-beam or flood irradiations.
• A new technique to utilize events with missing photosensor pixel values or timestamps to increase sensitivity.

Main Results:

• Achieved a spatial resolution of 1.7mm FWHM and a coincidence resolving time (CRT) of 214ps on a 32x32x22mm LYSO:Ce monolithic detector.
• Obtained an average DOI resolution of 3.7mm FWHM.
• Demonstrated the feasibility of using events with up to 16/64 missing photosensor pixels with minimal performance degradation.

Conclusions:

• The developed accelerated algorithms make monolithic scintillator detectors more practical for clinical TOF-PET applications.
• The revised methods simplify calibration and improve data processing efficiency.
• The new technique for handling incomplete data enhances system sensitivity without significant performance loss.