CMR in the Evaluation of Diastolic Dysfunction and Phenotyping of HFpEF: Current Role and Future Perspectives.

Heart failure with preserved ejection fraction presents a challenging diagnosis given a heterogeneous patient population and limited therapeutic options. Diastolic function assessment using echocardiography has been a cornerstone in the work-up and is as important as systolic functional assessment. There has been increased awareness to the potential utility of cardiac magnetic resonance (CMR) imaging over the past decade as a promising, radiation-free, robust imaging modality providing an unrestricted field of view and high-resolution images for global and regional functional assessment. CMR provides early markers for detecting myocardial disease using tissue characterization imaging, which might prove useful to improve diagnosis and management. Over the years, several studies have examined CMR-derived diastolic functional indices, including transmitral and pulmonary venous velocities, left ventricular and left atrial strain using myocardial tagging, and, more recently, feature tracking. The relevance of imaging-based diastolic function indices and their clinical application across different modalities is increasingly recognized.

Tissue volume and activity mapping using total intensity projection of PET/CT images.

Autoradiography using phosphor imaging screens is often used to characterize tissue distribution of positron emission tomography (PET) radiotracers. PET tracers emit positrons with limited penetration range, and valid quantitative autoradiography can therefore only be achieved in thin tissue slices. However, in some settings, quantitative tracer profiling in thick tissues is required. Our aim was to develop a reliable method for this purpose. In this paper, we present a method based on total intensity projections (TIPs) of PET and computed tomography (CT) images. We show theoretically and experimentally that tissue total activity and tissue volume maps can be derived from the TIPs of PET and CT images, respectively. We also show that these maps are free of signal displacement artifacts in the direction of projection. To demonstrate the utility of the approach, we obtain and compare TIP-based maps and autoradiography of ex-vivo atherosclerotic minipig aortas following in-vivo injection of 18F-fluorodeoxyglucose. We show that autoradiography of the thick aortas yields distorted results due to positron range effects, whereas TIP-mapping is free from such bias. The TIP-based maps may, thus, provide a low-resolution alternative to autoradiography, when tracer accumulation profiling in thick tissues is required.