18151497. TRAINING OF NEURAL NETWORK FOR ATTENUATION CORRECTION IN PET/CT simplified abstract (CANON MEDICAL SYSTEMS CORPORATION)

TRAINING OF NEURAL NETWORK FOR ATTENUATION CORRECTION IN PET/CT

Organization Name

CANON MEDICAL SYSTEMS CORPORATION

Inventor(s)

Yi Qiang of Vernon Hills IL (US)

Xiaohui Zhan of Vernon Hills IL (US)

Wenyuan Qi of Vernon Hills IL (US)

Jeffrey Kolthammer of Vernon Hills IL (US)

TRAINING OF NEURAL NETWORK FOR ATTENUATION CORRECTION IN PET/CT - A simplified explanation of the abstract

This abstract first appeared for US patent application 18151497 titled 'TRAINING OF NEURAL NETWORK FOR ATTENUATION CORRECTION IN PET/CT

The abstract describes a method for generating an attenuation map for PET image reconstruction using a deep convolutional neural network (DCNN) model trained with supervised data from spectral CT scans.

• Training a DCNN model to minimize the loss function between initial input image data and the attenuation map generated by a spectral CT scan.
• Obtaining PET data from a subject scan and reconstructing a PET image using the DCNN-generated attenuation map.
• Input image data can be from a conventional CT scan, a phantom, a simulation, or a SPECT image.

Potential Applications: - Improved accuracy and efficiency in PET image reconstruction. - Medical imaging for diagnosis and treatment planning. - Research in the field of deep learning and medical imaging technology.

Problems Solved: - Enhancing the quality of PET images by generating accurate attenuation maps. - Streamlining the process of PET image reconstruction. - Addressing challenges in medical imaging data processing.

Benefits: - Higher quality and more precise PET images. - Faster and more reliable image reconstruction. - Potential for improved diagnostic capabilities in healthcare.

Commercial Applications: Title: Advanced PET Image Reconstruction Technology for Medical Imaging Description: This technology can be utilized in hospitals, imaging centers, and research institutions for enhanced medical imaging services, leading to improved patient care and research outcomes.

Questions about PET Image Reconstruction Technology: 1. How does the use of a DCNN model improve the accuracy of attenuation map generation for PET image reconstruction? 2. What are the potential limitations or challenges associated with implementing this technology in clinical settings?

Frequently Updated Research: Stay updated on advancements in deep learning algorithms for medical imaging applications, as well as developments in PET image reconstruction techniques.

Original Abstract Submitted

A method is provided for generating an attenuation map for PET image reconstruction. The method includes training a deep convolutional neural network (DCNN) model by minimizing a loss function between initial input image data and the attenuation map generated by a spectral CT scan as supervised data. Further, the method includes obtaining PET data from a scan of a subject and reconstructing a PET image from the PET data and an attenuation map output from the DCNN. The initial input image data can be from a conventional CT scan with or without beam-hardening correction or the input image data can be from a phantom, a simulation, or a SPECT image.