%0 Journal Article %A Qingle Meng %A Feng Wang %A Yiming Du %A Runze Wu %A Chenwei Li %A Youjun Sun %A Tao Feng %T Consistency evaluation of myocardial viability and cardiac function between cardiac self-gating and ECG monitor based gating PET %D 2019 %J Journal of Nuclear Medicine %P 46-46 %V 60 %N supplement 1 %X 46Objectives: Cardiac studies using positron emission tomography (PET) require cardiac gating signals (CGS) which is usually obtained from an electrocardiography (ECG) recording device. A device-less cardiac gating technique called ‘self-gating’ can offer CGS in clinical PET system using the acquired list-mode data without attaching any hardware to patients. This work is a retrospective study to validate the consistency between cardiac self-gating and ECG monitor based gating used in myocardial viability studies. Methods: This retrospective study involved 3 patients who had gone through 15-minute single-bed list-mode ECG-gated 18F-FDG cardiac viability acquisition (uMI780, United Imaging Healthcare, Shanghai China). Both self-gated series and ECG-gated series were reconstructed as 8-gated frames with 2 iterations of OSEM, TOF + PSF, 256 x 256 image matrix and 3 mm of Gaussian smoothing. The quantitative results of both image sets were measured with Emory Cardiac Tool Box (ECTB), which included the viability results (defect mass and the defect ratio of each area), volume of the left ventricle (LV) of the patients in the end-of-systole phase (ESV) and end-of-diastole phase (EDV) and the ejection fraction (EF) of LV. The consistency of these results was evaluated with Bland-Altman plot. Results: All results were measured in pair from both self-gated image series and ECG monitor based gated image series. A total of 7 defection mass (gram) and 29 ratios of defective area (based on 17-segment left ventricle segmentation and measured as ratio) were measured in 3 patients. The Pearson correlation coefficient between two gating modes was 0.978 (P < 0.001) and 0.967 (P < 0.001) for defection mass and ratio of defective areas. The Bland-Altman plot presented that the mean of the difference between two gating modes were -1.29 ± 1.46 g and -1.07 ± 1.45%. The 95% limits of agreement (LoA) were (-8.85, 6.28) g and (-16.33, 14.19) %. As to cardiac functional measurement, the Pearson correlation coefficient of ESV, EDV and EF were 0.986 (P < 0.001), 0.966 (P < 0.001) and 0.954 (P < 0.001) respectively. The Bland-Altman plot showed that the mean differences were 6.00 ± 4.16 ml, 10.67 ± 2.96 ml and 1.67 ± 1.20% respectively. The 95% LoA were presented as (-20.13, 8.13) ml, (-20.72, -0.61) ml and (-5.75, 2.41) %. Conclusions: The new technology of cardiac self-gating system is able to provide a valid and accurate cardiac viability and function assessment using PET list-mode data compared with those using ECG monitor device. With the advantage of dispensing of hardware or end-user attachment, self-gating can be an ideal substitute for conventional ECG monitor based gating.Figure 1. The signal in orange and the signal in blue represent the ECG signal and the self-gated heart beat signal respectively and it shows good synchronism between ECG signal and self-gated signal.Figure 2. These are viability evaluation according to 17-segment left ventricle segmentation and the values represent the ratio of defect area in each segment. (A) was calculated with ECG signal gated data and (B) was calculated with self-gated data.Figure 3. These graphs are calculated with the ratios of defective areas acquired from ECG-gating and self-gating. The left plots the correlation between the two gating methods and the Pearson correlation efficient is 0.967. The right is the Bland-Altman plot of the two data sets and it shows 96.6% of the points (28 out of 29) are within the 95% LoA of (-16.33, 14.19) %U