Table 1 Comparative summary of diaCEST vs PET

Resolution

On the order of 1–2 mm

Depends on the hardware and pulse sequence, with no inherent limit. Generally a trade-off with acquisition time.

In the order of 4–5 mm

Nonuniform across the field-of-view

Intrinsically limited by positron range.

Sensitivity

Dependent on the targeted species as well as saturation scheme.

Generally micro- to millimolar concentrations, although nanomolar concentrations have been reported in some studies.

Can detect picomolar concentrations of radiotracer.

Selectivity

Selective for exogenous CEST agents.

Reduced selectivity for endogenous agents (several contributors, in addition to pH, temperature, buffer…)

Selectivity defined by radiotracer molecule.

Field-of-view

Inherently same as MR, in practice currently limited by the scan time.

Typically single or a few 2D slices.

3D acquisition is also feasible.

Up to whole-body acquisition, in blocks defined by the axial coverage of the detector, typically 15-25 cm.

Scan length

Typically 1–2 min per sweep of the spectrum.

Typically 2 min per bed position for body imaging, 10 min for brain.

Risks

SAR caused by excessive RFs may lead to heating damage. This is prevented by built-in software safety measures.

A dose of ionizing radiation is delivered to the patient, both by the radiotracer itself and by the transmission scan used for attenuation correction purposes.

Pitfalls

At present, only a limited number of metabolites can be detected at 3 T. CEST acquisitions are severely limited by motion, especially in areas of B0 or B1 inhomogeneity.

Reproducible CEST quantification is still under investigation.

Gamma attenuation information is required for quantitative reconstruction. This may be acquired by means of external gamma sources, or inferred from CT or MR images.

A significant delay is often required between radiotracer injection and imaging, for redistribution and uptake purposes.