Improved attenuation correction registration for FDG PET/CT images using data-driven gating (DDG)-based motion match

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiology

Background:

• Advancements in PET/CT imaging have highlighted the impact of respiratory motion artifacts on image quality.
• The motion-match CT (MMCT) algorithm aims to mitigate these artifacts by aligning CT data with PET data.
• MMCT-attenuation correction (MMCTAC) is proposed to improve whole-body FDG PET/CT reconstruction.

Purpose of the Study:

• To evaluate the effectiveness of MMCTAC in reducing respiratory motion artifacts in FDG PET/CT.
• To compare lesion detectability and image quality between MMCTAC and standard attenuation correction (normAC).

Main Methods:

• A cohort of 145 whole-body FDG PET/CT scans was analyzed, focusing on 23 high-motion patients with 36 small lesions.
• PET data were reconstructed using BSREM with ungated and gated data, employing both normAC and MMCTAC.
• An experienced radiologist performed qualitative analysis, scoring lesion detectability, diagnostic confidence, and image quality on a 5-point Likert scale.

Main Results:

• Clinicians ranked MMCTAC images higher than normAC images in 52% of cases.
• MMCTAC reconstructions showed positive mean differences in lesion detectability and diagnostic confidence compared to normAC.
• SUVmax values for lesions were significantly increased (p < 0.05) with MMCTAC across all reconstructions.

Conclusions:

• The MMCT algorithm effectively reduces respiratory motion artifacts in FDG PET/CT images.
• MMCTAC improves diagnostic capability for assessing lesions, particularly in the thorax and upper abdomen.
• This technique enhances overall image quality and diagnostic confidence in PET/CT imaging.