Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CT to PET Translation: A Large-scale Dataset and Domain-Knowledge-Guided Diffusion Approach

Dac Thai Nguyen, Trung Thanh Nguyen, Huu Tien Nguyen, Thanh Trung Nguyen, Huy Hieu Pham, Thanh Hung Nguyen, Thao Nguyen Truong, Phi Le Nguyen

TL;DR

A conditional diffusion model named CPDM is introduced, which is one of the initial attempts to employ a diffusion model for translating from CT to PET images, and surpasses existing methods in generating high-quality PET images in terms of multiple metrics.

Abstract

Positron Emission Tomography (PET) and Computed Tomography (CT) are essential for diagnosing, staging, and monitoring various diseases, particularly cancer. Despite their importance, the use of PET/CT systems is limited by the necessity for radioactive materials, the scarcity of PET scanners, and the high cost associated with PET imaging. In contrast, CT scanners are more widely available and significantly less expensive. In response to these challenges, our study addresses the issue of generating PET images from CT images, aiming to reduce both the medical examination cost and the associated health risks for patients. Our contributions are twofold: First, we introduce a conditional diffusion model named CPDM, which, to our knowledge, is one of the initial attempts to employ a diffusion model for translating from CT to PET images. Second, we provide the largest CT-PET dataset to date, comprising 2,028,628 paired CT-PET images, which facilitates the training and evaluation of CT-to-PET translation models. For the CPDM model, we incorporate domain knowledge to develop two conditional maps: the Attention map and the Attenuation map. The former helps the diffusion process focus on areas of interest, while the latter improves PET data correction and ensures accurate diagnostic information. Experimental evaluations across various benchmarks demonstrate that CPDM surpasses existing methods in generating high-quality PET images in terms of multiple metrics. The source code and data samples are available at https://github.com/thanhhff/CPDM.

CT to PET Translation: A Large-scale Dataset and Domain-Knowledge-Guided Diffusion Approach

TL;DR

A conditional diffusion model named CPDM is introduced, which is one of the initial attempts to employ a diffusion model for translating from CT to PET images, and surpasses existing methods in generating high-quality PET images in terms of multiple metrics.

Abstract

Positron Emission Tomography (PET) and Computed Tomography (CT) are essential for diagnosing, staging, and monitoring various diseases, particularly cancer. Despite their importance, the use of PET/CT systems is limited by the necessity for radioactive materials, the scarcity of PET scanners, and the high cost associated with PET imaging. In contrast, CT scanners are more widely available and significantly less expensive. In response to these challenges, our study addresses the issue of generating PET images from CT images, aiming to reduce both the medical examination cost and the associated health risks for patients. Our contributions are twofold: First, we introduce a conditional diffusion model named CPDM, which, to our knowledge, is one of the initial attempts to employ a diffusion model for translating from CT to PET images. Second, we provide the largest CT-PET dataset to date, comprising 2,028,628 paired CT-PET images, which facilitates the training and evaluation of CT-to-PET translation models. For the CPDM model, we incorporate domain knowledge to develop two conditional maps: the Attention map and the Attenuation map. The former helps the diffusion process focus on areas of interest, while the latter improves PET data correction and ensures accurate diagnostic information. Experimental evaluations across various benchmarks demonstrate that CPDM surpasses existing methods in generating high-quality PET images in terms of multiple metrics. The source code and data samples are available at https://github.com/thanhhff/CPDM.

Paper Structure

This paper contains 16 sections, 7 equations, 4 figures, 5 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. CT and PET Datasets
  4. Image-to-Image Translation
  5. Our Proposed PET/CT Dataset
  6. CPDM: Domain-Knowledge-Guided CT-to-PET Diffusion Model
  7. Motivation and Architecture Overview
  8. Domain Knowledge-Guided Maps
  9. Conditional Brownian Bridge Diffusion Process
  10. Performance Evaluation
  11. Experimental Settings
  12. Comparison with Existing Methods
  13. Ablation Studies
  14. Impacts of Domain Knowledge Guidance
  15. Robustness of the Condition Maps
  16. ...and 1 more sections

Figures (4)

  • Figure 1: Examples of synthetic PET images produced by natural image generative models (BBDM li2023bbdm, DDIB su2022dual, Palette saharia2022palette, and Pix2Pix wang2018highresolutionimagesynthesissemantic). The generated images do not accurately replicate the ground truth.
  • Figure 2: Overview of CPDM. The medical knowledge Attention map and the Attenuation map are used to guide each stage of the Brownian Bridge diffusion process.
  • Figure 3: Visualization of PET images generated by the best-performing methods. Errors induced by PET produced by CPDM are significantly more minor than other methods.
  • Figure 4: Visualization of spatial features extracted from the denoising network $\epsilon_\theta$.