Enhanced attenuation correction in FDG-PET/CT with respiratory mismatch correction
Yoshiyuki Kitamura1, Takuro Isoda2, Motohiro Sando2
1Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka-shi, Japan. kitamura.yoshiyuki.524@m.kyushu-u.ac.jp.
Background:
In positron emission tomography/computed tomography (PET/CT), CT is typically used for attenuation correction during the creation and reconstruction of PET images. Respiratory motion can cause artifacts due to misalignment of PET images, known as respiratory misalignment artifacts (RMA), which occur in up to 30% of lesion-based evaluations, often necessitating rescans of both PET and CT. A new method for reconstructing PET/CT artifacts caused by respiratory motion, termed 'Enhanced AC' (EAC) (GE HealthCare, USA), was developed to improve RMA. To evaluate the effectiveness of EAC, radiologists' visual assessment using a misalignment score is appropriate regardless of the presence or absence of target lesions. This study aimed to assess the usefulness of EAC based on radiologist evaluations.
Methods:
Consecutive F-18 fluorodeoxyglucose (FDG) PET/CT cases imaged at our hospital between July and August 2024 were included. Two radiologists independently assessed the deviation of FDG-PET/CT images from all cases imaged conventionally. Cases identified by either radiologist as misaligned were designated as the misalignment case group and used for evaluating EAC. A 5-point scoring system was established for the misalignment score. To evaluate the effect of EAC, PET coronal section images from the misalignment case group before EAC adaptation (pre-images) and the EAC-reconstructed images (EAC images) were displayed side by side. Five radiologists independently evaluated both images using a misalignment score.
Results:
Fifty-four patients (male/female = 36/18, average age 64.1 ± 15.0 years, median 66.5 years) were identified as the misalignment case group. The average misalignment scores from the five radiologists for the pre-images were 3.5, 3.4, 3.3, 3.1, and 3.1, respectively, whereas the scores for the EAC images were significantly lower (p < 0.01) across all radiologists: 2.1, 2.4, 2.5, 2.2, and 1.9, respectively.
Conclusions:
EAC is an effective and practical method for reducing RMA in selected cases. This method can be easily tailored to individual patients and does not require a large, deep-learning dataset, making it suitable for routine diagnostics.
