A multi-resolution TOF-DOI detector for human brain dedicated PET scanner

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Detector Physics

Background:

• Positron Emission Tomography (PET) scanners require high resolution for effective human brain imaging.
• Existing PET detector designs face limitations in simultaneously achieving high spatial, depth-of-interaction (DOI), and time-of-flight (TOF) resolutions.
• Advanced detector designs are crucial for next-generation brain-dedicated PET systems.

Purpose of the Study:

• To propose and evaluate a novel multi-resolution detector design for brain-dedicated PET scanners.
• To achieve high spatial, DOI, and TOF resolutions with high sensitivity.
• To assess the performance of a two-layer detector system for enhanced PET imaging.

Main Methods:

• A two-layer detector design featuring LYSO crystal arrays and a lightguide was developed.
• The top layer used smaller crystals for spatial resolution, while the bottom layer used larger crystals for TOF resolution.
• A convolutional neural network was employed for layer discrimination, and inter-crystal scatter (ICS) events were analyzed.

Main Results:

• The convolutional neural network accurately distinguished between top and bottom detector layers with 97% accuracy.
• High spatial resolution was achieved by the top layer, and high TOF resolution was achieved by the bottom layer (277 ps after ICS exclusion).
• The detector demonstrated a measured average DOI resolution of 4.1 mm and high detection efficiency.

Conclusions:

• The proposed two-layer detector design successfully integrates high spatial and TOF resolutions for brain PET.
• The detector's high DOI resolution and sensitivity make it suitable for advanced brain imaging applications.
• This design represents a significant advancement for next-generation high-performance brain-dedicated PET scanners.