Generation of 4-dimensional CT images based on 4-dimensional PET-derived motion fields

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiology

Background:

• Respiratory motion degrades accuracy in multimodality image fusion, blurring lesions and underestimating uptake.
• Current solutions like respiration-synchronized acquisitions improve image quality but may have discrepancies between PET and CT data.
• Existing methods often rely on 4D CT, increasing patient radiation dose and potentially introducing acquisition-related variations.

Purpose of the Study:

• To develop a method for generating dynamic CT images from 4D PET data, eliminating the need for dynamic CT acquisition.
• To improve the accuracy of attenuation correction in respiration-synchronized PET scans by ensuring better CT-PET image matching.
• To reduce radiation dose burden associated with 4D CT in multimodality imaging.

Main Methods:

• A novel approach generates dynamic CT images by combining a reference CT image with deformation matrices derived from elastic registration of 4D PET images.
• This method utilizes 4D PET images that are not corrected for attenuation.
• The technique was validated using Monte Carlo simulations and patient datasets, assessing the impact of PET image noise.

Main Results:

• The proposed method successfully generated dynamic CT images, demonstrating good agreement with original CT images in patient studies.
• The approach is robust across different levels of PET image statistical quality, indicating resilience to image noise.
• Validation confirmed the feasibility of using clinically relevant PET acquisition times.

Conclusions:

• The developed method provides an effective way to generate dynamic CT images without additional radiation exposure.
• Accurate CT-PET image alignment is achieved, enabling precise attenuation correction for synchronized PET acquisitions.
• This approach shows promise for application in PET/CT and potentially other multimodality systems like PET/MR.