Skip to main content
2025 SCMR Conference Main banner
Final Program


Schedule at a Glance
Back

Rapid Fire Abstracts

Rapid Fire 02- Multiparametric Mapping I

Impact of Magnetization Transfer Effects on T1 and T2 Measurements using Cardiac Magnetic Resonance Fingerprinting (RF_TH_192)

Thursday, January 30, 2025
2:20 PM – 2:30 PM East Coast USA Time
Location: Blue Room PreFunction
Background: Myocardial tissue characterization with cardiac magnetic resonance is an essential tool for detecting and differentiating between various myocardial diseases. In particular, elevations in myocardial T1 and T2 values can indicate fibrosis, myocarditis, infarction, and oedema among others, making accurate tissue property mapping crucial for disease diagnosis1–4. T1 and T2  can be mapped by multiple different methods, including simultaneously using Cardiac Magnetic Resonance Fingerprinting (cMRF). All of these mapping techniques, including cMRF, work by collecting weighted images, and then fitting individual pixel values to a signal model to obtain maps; the models used for fitting often ignore effects such as Magnetization Transfer (MT) in order to simplify the fitting process. However, neglecting MT may reduce the accuracy of tissue property measurements5. For example, MT has been shown to influence Magnetic Resonance Fingerprinting (MRF) measurements in the brain6. In this study, the effects of MT on T1 and T2 measurements made with cMRF are explored by generating maps both with and without including MT in the signal modeling. The effects of ignoring and including MT in cMRF are evaluated in the myocardium of healthy subjects at 1.5T.

Methods: T1 and T2 maps were acquired in a single, mid-ventricular short-axis slice in 14 healthy subjects and 1 cardiomyopathy patient at 1.5T (MAGNETOM Sola, Siemens Healthineers) using a 2D, ECG-triggered breath-held cMRF sequence (1.6x1.6x8mm3, FA 4-25°, constant TR/TE 5.4/1ms)7. Note that no off-resonance MT preparation pulses were included in the sequence. LGE images were acquired in the patient ~10min after injection of contrast. T1 and T2 maps were reconstructed using a deep image prior method8 that includes conventional signal modeling (i.e. ignoring MT effects) and signal modeling that includes MT effects using the EPG-X framework9. The EPG-X framework models exchange between free-water and bound-water pools assuming a fixed rate of MT between pools (kf=4.2Hz)10, T1,f=T1,b, and T2,b=12μs, which enables T1,f, T2,f, and the size of the bound pool (referred to as the bound pool fraction) to be mapped. T1 and T2 values (or T1,f, T2,f, and bound pool fraction when MT is included) were obtained by computing the voxelwise mean and standard deviation from ROIs manually drawn within the six mid ring American Heart Association (AHA)11 segments on the maps. Paired t-test between myocardial tissue property values measured with and without modeling MT effects were performed.

Results: Representative T1/T1,f, T2/T2f, and bound pool fraction maps obtained from a healthy volunteer reconstructed both ignoring and including MT effects in the signal model are shown in Fig. 1. Fig. 2 shows average T1 and T2 values measured in each of the six AHA segments for both types of reconstructions. Significant increases in both T1 and T2 values (p< 0.001) were observed when accounting for MT in the signal model. Fig. 3 shows MT-modeled native T1, T2, and bound pool fraction maps in a patient along with the corresponding LGE image. Note the tissue property value changes in the maps corresponding to regions of enhancement on LGE.

Conclusion: MT influences the tissue property values measured using cMRF and MT must be taken into account in the signal model to produce accurate measurements of T1 and T2. The addition of the bound pool fraction map achieved by modeling MT effects may provide additional insight into the macromolecular make-up of the myocardium which could potentially aid diagnosis.