Deep Learning Attenuation Correction for Cardiac SPECT Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional attenuation correction techniques for SPECT myocardial perfusion imaging require dedicated SPECT/CT scanners, leading to limited availability and increased radiation exposure, and are susceptible to misregistration errors, which diminish the clinical value of the images.
Innovation Solution
A deep learning-based method using a conditional generative adversarial network (cGAN) generates simulated attenuation-corrected SPECT imaging data from non-corrected data, eliminating the need for separate CT imaging and reducing radiation exposure by applying attenuation correction directly to SPECT images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional attenuation correction techniques using separate CT scans are used, then diagnostic accuracy is improved, but radiation exposure increases
Solution Approach 1:
The patent creates a simulated CT attenuation map by copying and adapting the existing SPECT imaging data through deep learning processing. Instead of acquiring separate CT data, the system generates a synthetic attenuation correction map from the SPECT images themselves, thereby achieving attenuation correction without additional radiation exposure from CT scans
Solution Approach 2:
The patent replaces the physical CT scanning mechanism with a computational deep learning model. The cGAN-based system substitutes the mechanical x-ray based CT acquisition process with an artificial intelligence-based image processing approach that processes existing SPECT data to generate attenuation correction maps
2Measurement precision
If dedicated SPECT/CT scanners are used for attenuation correction, then diagnostic accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The patent makes the SPECT scanner perform multiple functions by enabling it to generate attenuation correction maps in addition to its primary imaging function. The deep learning model allows the SPECT system to derive CT-like attenuation information from its own imaging data, eliminating the need for separate CT hardware and making the SPECT scanner a multi-functional device
Solution Approach 2:
The patent enables the SPECT imaging system to correct its own images by generating attenuation correction maps from its own acquired data. The cGAN model processes the SPECT images to create the necessary correction information, allowing the system to be self-sufficient and eliminating dependency on external CT scanners
3Measurement precision
If separate CT and SPECT scans are acquired for attenuation correction, then correction accuracy is improved, but misregistration errors increase
Solution Approach 1:
The patent merges the attenuation correction process with the SPECT imaging process by generating the attenuation map directly from the SPECT data. This integration eliminates the separate acquisition step and the associated registration problem, as the attenuation correction is derived from the same imaging session and coordinate system
Data Source
AI summary
Systems and methods are disclosed for applying attenuation correction to single photon emission computed tomography (SPECT) imaging data for myocardial perfusion imaging (MPI) studies. SPECT-MPI imaging data can be provided to a deep-learning model to automatically generate simulated computed tomography attenuation correction (CT-AC) images from the non-corrected (NC) SPECT-MPI imaging data. These simulated CT-AC images can then be used to perform attenuation correction on the SPECT-MPI imaging data to generate corrected SPECT-MPI imaging data. The deep-learning model can be trained using corresponding pairs of non-corrected SPECT-MPI imaging data and traditional CT-AC imaging data. The deep-learning model can be a conditional generative adversarial neural network (cGAN).