Evaluation of Single-Chip, Real-Time Tomographic Data Processing on FPGA SoC Devices.

Area of Science:

• Medical Imaging
• Computer Engineering

Background:

• Traditional tomographic data processing requires substantial local computing power for real-time image reconstruction.
• Existing systems face challenges in efficiently handling large data volumes generated by modern scanners.

Purpose of the Study:

• To develop and evaluate a novel, decentralized approach for tomographic data processing.
• To reduce the reliance on powerful, on-site processing facilities for positron emission tomography (PET) scanners.
• To streamline the data processing pipeline from raw scanner output to visualization.

Main Methods:

• Implementation of a field-programmable gate array (FPGA) system-on-chip (SoC) platform directly interfaced with PET scanner data streams.
• Utilizing programmable logic for core processing tasks: event building, filtering, coincidence detection, and region-of-response (RoR) reconstruction.
• Employing integrated processors for on-the-fly visualization.

Main Results:

• The proposed FPGA-SoC platform successfully performed event building, filtering, coincidence search, and RoR reconstruction.
• Significant reduction in data volume was achieved by converting raw data to a list-mode representation.
• On-the-fly visualization was successfully generated, demonstrating real-time processing capabilities.

Conclusions:

• The developed FPGA-SoC approach offers an efficient and scalable solution for tomographic data processing.
• This novel system eliminates the need for powerful local processing facilities, making advanced PET imaging more accessible.
• The method effectively reduces data volume and provides real-time visualization, enhancing the usability of PET scanners.