Using PET to determine the atherosclerotic burden in the cardiovascular system

Introduction: According to the WHO, ischemic heart disease and stroke are the world’s number one and two killers. Atherosclerosis is usually the main cause, which may be asymptomatic and usually diagnosed with a complication. Atherosclerosis may be discovered during a health check or using the most common types of imaging modalities which can determine if a stenosis is present. Early changes in the artery wall can be detected and measured by Positron emission tomography (PET) imaging with tracers 18F-fluorodeoxyglucose (FDG) or Sodium Fluoride (NaF). So, it was initially tried to systemically review literature about relationship between PET findings and other clinical and paraclinical aspects in carotid arteries. Second, we investigated NaF uptake in the major arteries of a healthy and a high-risk group during two years of follow-up in a cohort study. Finally, we designed an artificial intelligence (AI)-based model to segment the heart and aorta, then compared the results with the manual segmentation.

Methods: In the systematic review, research articles about carotid artery PET imaging with different radiotracers were searched in several databases. In the cohort study, 29 healthy subjects and 20 angina pectoris patients underwent NaF-PET/CT twice two years apart. The arch of aorta, thoracic aorta, abdominal aorta and carotids were manually segmented. NaF uptake was expressed as the maximum, mean and total standardized uptake values without and with partial volume correction (SUVmax, SUVmean, SUVtotal and cSUVmean, cSUVtotal). A convolutional neural network (CNN) based method was developed to identify and segment the heart and aorta in three sections: the arch, thoracic and abdominal aorta. This CNN model was tested in NaF- PET/CT scans of 49 subjects, then compared with data obtained by manual segmentation.

Results: FDG-PET visualizes the inflammatory part of carotid atherosclerosis enabling risk stratification to a certain degree, whereas NaF-PET seems to indicate long-term consequences of ongoing inflammation by demonstrating microcalcification. In the cohort study, NaF uptake was insignificantly higher in the angina group at both time points, with less uptake in the healthy group and higher in the angina group after two years. In the final part of the project, CNN-derived heart segmentation measures were 0% to 4% higher than the manual method and 0% to 17% lower in the aortic segmentation. However, the CNN and manual methods of SUVmean values in both heart and aortic segmentation were almost identical. The heart and aortic CNN-based segmentation method was faster than the manual method, which had maximal 0.5% and 6% variation in repeated manual segmentation in heart and aortic segmentation, respectively.

Conclusion: PET imaging is a newly introduced modality for imaging of atherosclerotic changes, which is a slow and variable process among individuals with different medical backgrounds. albeit with a tendency of slight NaF uptake increases in angina patients. Although current technical difficulties such as time-taking image analysis exist, the AI-based models could present values for Volume, SUVmean, SUVmax, and SUVtotal similar to the manually obtained ones. These AI-based models were more reproducible and quicker alternatives for slow manual segmentation.

Funding:

• University of Southern Denmark (Faculty scholarship)
• The Region of Southern Denmark (The regional fundig to support highly specialized services)