Singles Pairing for PET Imaging Using Software Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional PET systems face limitations in flexibility, hardware reuse, cost, and efficiency due to rigid hardware implementations for singles pairing, which restricts transaxial imaging FOV and requires separate coding for different scanner geometries and variants, leading to increased engineering and material costs.
Innovation Solution
A software-based method for processing PET information using a processor to pair singles events into coincidences, allowing for flexible pairing schemes and accepting or rejecting multi-coincidence events based on adjustable parameters, implemented using FPGA for on-line processing and CPU for off-line processing, enabling precise FOV selection and hardware reuse across different systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If hardware implementation of pairing using and-logic and/or ASIC chips is used, then pairing function is achieved, but flexibility is reduced and development cost increases
Solution Approach 1:
The patent replaces hardware-based pairing (using and-logic and ASIC chips) with software-based pairing implemented on general-purpose processors. This substitution eliminates the need for specialized hardware circuits, thereby increasing flexibility and reducing hardware complexity while maintaining the pairing function.
Solution Approach 2:
The patent employs a universal software-based pairing approach that can be applied across different scanner geometries and variants without requiring specialized hardware for each configuration. This universal software implementation allows the same system to adapt to multiple applications and geometries, enhancing versatility.
2Reliability
If specialized hardware (and-logic and/or ASIC chips) for pairing is used, then pairing function is achieved, but cost increases due to re-spin requirements
Solution Approach 1:
The patent replaces specialized hardware pairing with software-based pairing, eliminating the need for expensive ASIC fabrication and re-spin operations. Software can be updated and modified without manufacturing costs, significantly reducing development and maintenance expenses while maintaining reliable pairing functionality.
3Area of stationary object
If detector block/module pair defines transaxial imaging FOV, then coarse FOV determination is achieved, but precise FOV selection requires additional processing
Solution Approach 1:
The patent performs preliminary FOV selection during the pairing process itself rather than requiring separate post-processing steps. The software-based pairing method incorporates FOV determination logic that can precisely select the transaxial imaging FOV based on the paired events, eliminating the need for additional trimming operations.
4Speed
If hardware pairing is used, then pairing speed is achieved, but hardware reuse between different systems is limited
Solution Approach 1:
The patent replaces dedicated hardware pairing circuits with software-based pairing that can run on general-purpose processors. This software implementation can be easily transferred and adapted between different scanner systems and geometries, enabling full hardware reuse while maintaining pairing performance through efficient algorithm implementation.
Data Source
AI summary
A method of processing, by a processor of a computer, positron emission tomography (PET) information obtained from a PET detector, including the steps of obtaining a singles event list that includes a plurality of single event entries, each entry corresponding to a single event detected by the PET detector and including a timestamp, energy information, and crystal position information; and determining, based on the timestamp and the crystal position information of each entry and a predetermined coincidence time period, all event pairs within the plurality of single event entries, each event pair consisting of two single events whose respective timestamps differ by less than the predetermined coincidence time period, wherein the determining step determines that an event pair exists when at least two, but no more than k, single events are found within a given time window, wherein k is an integer value greater than or equal to two.


