Feasibility of finger-pulse triggering as a substitute for electrocardiogram triggering in cardiac magnetic resonance T1 mapping: a prospective comparative study

Abstract

Background

Electrocardiogram (ECG) triggering in cardiac magnetic resonance (MR) faces limitations with high-field systems (≥3.0T) due to magnetohydrodynamic (MHD) artifacts. This study aimed to evaluate the feasibility of using finger-pulse triggering for cardiac MR T1 mapping by comparing image quality and native T1 quantification with conventional ECG triggering.

Methods

Patients who underwent cardiac MR examination between March and April 2024 were prospectively and consecutively enrolled in the study. All the patients underwent identical pre-contrast T1 mapping with both ECG- and pulse-triggering acquisitions using a modified Look-Locker inversion-recovery (MOLLI) 5(3)3 sequence, covering short-axis views at basal, mid, and apical slices of the left ventricle. Three radiologists independently evaluated image quality using the Likert scale (range, 1–5). Two radiologists manually delineated myocardial regions of interest (ROIs) on native T1 maps to quantify segmental native T1 values. Paired t-tests or Wilcoxon signed-rank tests were used to compare the characteristics derived from the two triggering images. Bland-Altman plots, Kendall’s W test, and intraclass correlation coefficients (ICCs) were used for the agreement analysis.

Results

A total of 15 participants were included in the study (mean age: 41±19 years; 60% male). ECG and finger-pulse triggering demonstrated equivalent diagnostic image quality (median scores: 4.0 vs. 4.0, P=0.655), with excellent interobserver agreement (ECG: Kendall’s W =0.861, P<0.001; pulse: Kendall’s W =0.838, P=0.001). There were no statistically significant differences in the native T1 values between the two methods (all P>0.05). Bland-Altman plots revealed that the differences between the two triggering methods for native T1 values mostly fell within the 95% confidence interval. Both triggering modalities demonstrated good to excellent interobserver and intraobserver agreement (ICC range, 0.768–0.936).

Conclusions

Finger-pulse triggering demonstrates comparable accuracy and reliability to ECG triggering in cardiac MR T1 mapping, offering a viable clinical alternative for patients with ECG distortion or triggering failure.

Introduction

Cardiac magnetic resonance (MR) has emerged as the gold standard non-invasive imaging tool for comprehensively evaluating cardiac anatomy, function, and tissue characteristics (1). Advanced myocardial mapping techniques, particularly T1 mapping, enable quantitative tissue characterization through the analysis of T1 relaxation time variations, facilitating the detection of fibrosis, edema, amyloidosis, and lipid accumulation (2). This capability positions T1 mapping as a pivotal tool for early diagnosis and precision medicine in cardiovascular disease.

T1 mapping relies on co-registered images acquired at sequential time points during T1 recovery (3). Most protocols employ balanced steady-state free precession sequences, with the modified Look-Locker inversion-recovery (MOLLI) technique being the predominant method for myocardial T1 quantification (4,5). Recent advancements, including improved MOLLI, saturation recovery single-shot acquisition, along with their variants, have further enhanced clinical applicability of T1 mapping (5). Notably, the utility of T1 mapping extends beyond myocardial assessment. For example, cardiac MR-derived liver T1 mapping has been shown to be significantly correlated with right ventricular function and myocardial fibrosis in dilated cardiomyopathy, and thus has potential prognostic value (6).

Despite its diagnostic value, cardiac MR faces inherent challenges related to cardiac and respiratory motion. Current cardiac gating predominantly uses electrocardiogram (ECG) R-wave triggering to synchronize data acquisition with the cardiac cycle (7). However, ECG signals are susceptible to interference from three primary sources: gradient interference, radiofrequency (RF) interference, and the magnetohydrodynamic (MHD) effect (8,9). Gradient switching and changes in RF pulses induce voltage fluctuations in ECG leads, which distort ECG signals (8,10). Many studies have developed methods to generate synchronous signals to eliminate these artifacts (11,12). One study modeled gradient-induced interference as a linear time-invariant system and estimated its impulse response to precisely remove artifacts without signal loss (11). Another study employed parallel transmission with multi-channel coils arrays, incorporating RF shimming, k-space subsampling, self-calibration, and specific absorption rate constraints to optimize RF pulses. This strategy improved B₁ (RF field) homogeneity, reduced system-related errors, and effectively suppressed RF-induced artifacts (12).

The MHD effect arises from the interaction between conductive blood flow and the static magnetic field (B₀), generating voltages and currents perpendicular to both B₀ and flow direction, which are superimposed on the ECG signal (13). This phenomenon is most pronounced during early systole, when rapid ventricular ejection directs blood into the aortic arch (14). The MHD-induced currents predominantly distort T-wave morphology, often causing T-wave widening or elevation exceeding the amplitude of the R wave (8,15). These alterations lead to the false identification of T waves as R waves, compromising acquisition triggering accuracy in cardiac MR (9). The magnitude of the MHD effect is strongly correlated with the magnetic field strength. Significant ECG distortions have been documented at 3T (16), with these effects becoming more pronounced in ultra-high field systems (8). Current mitigation strategies include vector ECG-based triggering (17) and independent component analysis algorithms (14). However, technical limitations and implementation complexity have hindered their widespread clinical adoption.

When ECG triggering fails, peripheral pulse triggering is a valuable alternative. Photoplethysmography (PPG)-based systems detect blood volume changes via optical absorption/scattering in fingertip capillaries (18). The PPG waveform’s systolic upslope is correlated with ventricular contraction, while its diastolic decay reflects relaxation, enabling cardiac cycle synchronization (19). Moreover, the wave peak of the finger pulse is relatively broad and minimally affected by the magnetic fields, resulting in improved stability (19). Clinical validation demonstrates that finger-pulse triggering achieves comparable accuracy to ECG triggering for left ventricular (LV) functional assessment in 3T cardiac MR (20), suggesting its potential as a useful alternative triggering modality.

Despite these advancements, no prior investigations have evaluated the feasibility of finger-pulse triggering for T1 mapping. This study conducted a head-to-head comparison of pre-contrast T1 mapping using ECG and finger-pulse triggering to evaluate the technical feasibility of pulse-triggering T1 mapping, assess the agreement between the two triggering methods, and explore clinical scenarios suitable for pulse triggering. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-900/rc).

Methods

Study population

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Tongji Hospital (IRB No. TJ-IRB20230914). Written informed consent was obtained from all participants prior to enrollment.

A total of 18 consecutive patients were prospectively enrolled in the study between March and April 2024. The inclusion criteria were as follows: (I) a clinical indication for cardiac MR examination; (II) an adequate breath-holding capacity (i.e., an ability to follow breathing instructions and maintain a stable breath-hold for ≥15 seconds); and (III) the successful acquisition of both ECG- and pulse-triggered pre-contrast T1 maps. The exclusion criteria were as follows: (I) arrhythmia (e.g., atrial fibrillation) or a rapid heart rate (>100 beats/min); and/or (II) a peripheral circulation disorder. After the inclusion and exclusion criteria were applied, a total of 15 patients were included in the study.

Cardiac MR image acquisition

All patients underwent cardiac MR examination on a 3.0T magnetic resonance scanner (MAGNETOM Skyra, Siemens Healthcare, Erlangen, Germany), with an 18-channel body coil, a 32-channel spine-array coil, and prospective ECG gating and finger-pulse triggering. Before the scan, all the patients were instructed to hold their breath after exhalation as directed. Scanning was performed in the head-advanced supine position. Four magnetic resonance imaging (MRI)-compatible electrodes were placed according to the manufacturer’s recommended positions. ECG signals from leads I, II, and III were acquired via a non-magnetic amplifier, and the lead with the highest R-wave amplitude, lowest T-wave amplitude, and fewest artifacts was selected to trigger cardiac MR data acquisition. The finger pulse clip was placed on the left or right index finger, and triggered based on the highest pulse wave. The pulse-triggered acquisition compensated for pulse transit time (PTT) delay through the real-time adjustment of the acquisition window, applying a 200–250 ms temporal offset relative to the ECG R wave to ensure diastolic phase sampling. When the other parameters were fixed, the same slice localization was selected, and the ECG trigger and pulse trigger were performed in the basal, mid, and apical slices, and heart rate and respiratory stability were monitored during the process.

Based on a MOLLI sequence, pre-contrast T1 mapping images were acquired at LV short-axis basal, mid, and apical slices using a 5(3)3 acquisition protocol (21). Under this two-inversion protocol, images were acquired for 5 heartbeats following the first inversion, after which there was a 3-second recovery period, and images were then acquired for 3 heartbeats following the second inversion. Typical scanning parameters were as follows: slice thickness =5 mm, slice spacing =2 mm, repetition time (TR) =3.2 ms, echo time =1.2 ms, flip angle =35°, field of view (FOV) =360×360 mm², FOV phase =100%, matrix size =256×192, base resolution =256, phase resolution =75%.

Image quality evaluation

The T1 mapping images were independently assessed by three radiologists (D.T., with 10 years of experience; Z.Y., with 7 years of experience; Z.P., with 2 years of experience), who were blinded to the patient clinical information, diagnosis, and acquisition methods (i.e., ECG triggering and finger-pulse triggering). All images were presented in randomized order and assessed using a Syngo MMWP VE40B workstation (MAGNETOM Skyra, Siemens Healthcare). The ordinal scores of the three radiologists were averaged.

Subjective general image quality was rated on a five-point Likert scale as follows: 5= excellent image quality, interpretable with minimal to no artifacts; 4= good image quality, interpretable with mild artifacts; 3= adequate image quality, interpretable with moderate artifacts; 2= fair image quality, interpretable with obvious artifacts; and 1= poor image quality, uninterpretable images (22).

Native T1 quantification

All the cardiac MR T1 mapping images were analyzed by two experienced radiologists blinded to the patients’ clinical history using CVI42 (version 5.14.0, Circle Cardiovascular Imaging, Calgary, Canada), and reviewed by one observer twice at an interval of 2 weeks to avoid bias. Following the American Heart Association (AHA) 17-segment model (23), we adopted a modified 16-segment approach that excluded the apical cap for LV segmentation. The radiologists performed semi-automatic contouring of the endocardial and epicardial borders, with manual adjustments as needed, followed by the manual placement of region of interest (ROI) markers for each segment. The myocardial ROI was delineated as far as possible to avoid the endocardium and epicardium, and the location of the measured segments and the size of the ROI were consistent. Notably, ROIs smaller than 20 pixels were avoided to reduce partial volume effects, and ensure measurement accuracy and reproducibility (1).

Statistical analysis

All the data analyses were performed using SPSS (version 29.0.1.0; IBM, Armonk, NY, USA). Graphs were generated using GraphPad Prism (version 10.2.0; GraphPad Software, San Diego, CA, USA) and MATLAB (version R2021b; MathWorks, Natick, MA, USA). The Shapiro-Wilk test was used to assess the normality of the continuous variables. The normally distributed variables are presented as the mean ± standard deviation, and were compared using the paired t-test. The non-normally distributed variables are presented as the median with interquartile range (IQR), and were compared using the Wilcoxon signed-rank test. The categorical variables are presented as the absolute number and percentage. A Bland-Altman analysis was performed to evaluate the agreements and differences of native T1 values of different triggering methods. Interobserver agreements in the subjective general image quality assessment were evaluated with Kendall’s W, which was categorized as poor (<0.40), fair (0.25-0.60), good (0.61–0.80), or excellent (>0.80). Intra- and interobserver agreements in native T1 mapping quantification were evaluated using the intraclass correlation coefficient (ICC), and were labeled as poor (<0.50), fair (0.50–0.75), good (0.76–0.90), or excellent (>0.90). A two-tailed P value <0.05 was considered statistically significant.

Results

Participant characteristics

A total of 18 participants met the inclusion criteria; however, three patients were excluded because of arrhythmia (n=1), rapid heart rate (>100 bpm, n=1), and peripheral circulation disorders (n=1), respectively. Thus, the final study sample comprised 15 participants (mean age: 41±19 years; 60% male). The clinical diagnoses of the participants included hypertrophic cardiomyopathy (n=6), myocarditis (n=3), systemic disease cardiac involvement (n=2), hypertensive heart disease (n=1), cardiac tumor (n=1), and no clinical finding reported (n=2). The demographic characteristics of the patients were collected from clinical records and are summarized in Table 1.

Table 1. Patients’ clinical characteristics (n=15).

Characteristics	Values
Male	9 [60]
Age (years)	41±19 (10–66)
Height (cm)	162.6±8.3 (140–175)
Weight (kg)	63.8±12.9 (41–82)
Body surface area (m2)	1.69±0.21 (1.26–2.00)
Heart rate (bpm)	74.6±11.5 (55–92)
Diagnosis
Hypertrophic cardiomyopathy	6
Myocarditis	3
Cardiac involvement of systemic disease	2
Hypertensive heart disease	1
Cardiac tumor	1
Negative finding	2

Continuous normally distributed data are expressed as the mean ± standard deviation (range), and categorical data are expressed as the n [percentage] or n.

Image quality evaluation

The T1 mapping sequences were acquired using MOLLI 5(3)3 mode. ECG triggering and finger-pulse triggering showed comparable image quality, with identical median Likert scores [4.0 (IQR, 3.7–4.7) vs. 4.0 (IQR, 3.8–4.7), P=0.655], indicating no significant difference between the two methods. Representative T1 mapping images of ECG triggering and finger-pulse triggering are shown in Figure 1.

Figure 1.

Image examples of pre-contrast T1 mapping images acquired by ECG triggering and pulse triggering in three short-axis slices (basal, mid, and apical). ECG, electrocardiogram.

Native T1 quantification

There was no significant difference in the native T1 quantification between the ECG-triggering and pulse-triggering imaging of the 16 segments (Table 2) (all P>0.05). The mean native T1 values of the participants in three short-axis slices are displayed as bullseye plots in Figure 2, and the quantification results between the ECG triggering and finger-pulse triggering in all 16 myocardial segments were similar. Figure 3 shows the comparable native T1 values between both triggering methods through boxplot visualization, with the statistical analysis confirming no significant difference (P=0.081). The Bland-Altman scatter plots (Figure 4) illustrated that the differences between the two triggering methods for native T1 values mostly fell within the 95% confidence interval. The differences in basal anteroseptal (bias =0.27 ms) and basal anterolateral (bias =0.53 ms) were closer to zero compared to those in the other segments.

Table 2. Image quality assessment and native T1 quantification between ECG-triggering and finger pulse-triggering cardiac MR T1 mapping.

Parameter	ECG triggering	Pulse triggering	P value
General image quality	4.0 (3.7, 4.7)	4.0 (3.8, 4.7)	0.655
Native T1 value quantification
Basal
Anterior	1,269.87±28.61	1,271.40±27.48	0.474
Anteroseptal	1,290.40±36.04	1,290.13±31.56	0.911
Inferoseptal	1,285.93±33.83	1,290.53±28.23	0.236
Inferior	1,274.40±31.40	1,279.07±29.41	0.112
Inferolateral	1,271.27±35.28	1,269.13±37.95	0.519
Anterolateral	1,262.33±19.22	1,261.80±32.45	0.934
Mid
Anterior	1,269.20±43.07	1,271.53±36.96	0.637
Anteroseptal	1,302.73±37.47	1,305.93±38.97	0.258
Inferoseptal	1,296.67±28.67	1,295.27±29.98	0.590
Inferior	1,262.20±36.94	1,265.53±38.36	0.332
Inferolateral	1,250.93±29.96	1,254.53±34.39	0.310
Anterolateral	1,283.53±32.52	1,282.27±30.83	0.401
Apical
Anterior	1,271.40±45.39	1,273.93±28.47	0.465
Septal	1,291.13±28.47	1,292.33±30.26	0.531
Inferior	1,279.27±25.71	1,280.67±28.13	0.543
Lateral	1,278.47±42.85	1,279.47±38.43	0.652

Continuous normally distributed data are expressed as median (Q1, Q3) or mean± standard deviation. ECG, electrocardiogram; MR, magnetic resonance.

Figure 2.

The mean native T1 values of participants in three short-axis slices are displayed as bullseye plots (AHA-16-segment-model). AHA, American Heart Association; ECG, electrocardiogram.

Figure 3.

Comparison of native T1 values between ECG- and pulse-triggering acquisitions. The boxplots display the median (central line), interquartile range (box), and full data range (whiskers) from end-diastolic measurements. ns: (P=0.081). ECG, electrocardiogram; ns, not significant.

Figure 4.

Bland-Altman plots for native T1 values of 16 segments showing agreement between ECG triggering and pulse triggering. The red solid lines represent the average difference, and the black dotted lines represent the 95% confidence interval. ECG, electrocardiogram; SD, standard deviation.

Agreements of scoring and quantification

The interobserver agreement of the general image quality evaluation of the two triggering methods was excellent (ECG: Kendall’s W =0.861; pulse: Kendall’s W =0.838) (Table 3). The native T1 value of ECG triggering and pulse triggering demonstrated good to excellent inter- and intraobserver agreement (all ICC values >0.75). The detailed inter- and intraobserver agreements of the native T1 quantification results are shown in Table 4.

Table 3. Interobserver agreement for the image general quality scores for ECG and pulse triggering.

Group	Image quality scores	Kendall’s W	P value
ECG triggering		0.861	<0.001
Observer A	4 (3, 5)
Observer B	4 (4, 5)
Observer C	4 (4, 5)
Pulse triggering		0.838	0.001
Observer A	4 (3, 5)
Observer B	4 (4, 5)
Observer C	4 (4, 5)

Continuous non-normally distributed data are expressed as median (Q1, Q3). ECG, electrocardiogram.

Table 4. Intra- and interobserver reproducibility for native T1 measurements assessed by ICCs with 95% confidence intervals.

Parameter	ECG triggering			Pulse triggering
	Intraobserver	Interobserver		Intraobserver	Interobserver
Basal
Anterior	0.919 (0.777–0.972)	0.929 (0.803–0.976)		0.860 (0.633–0.951)	0.845 (0.600–0.945)
Anteroseptal	0.903 (0.735–0.966)	0.918 (0.774–0.972)		0.909 (0.751–0.968)	0.900 (0.730–0.965)
Inferoseptal	0.936 (0.820–0.978)	0.934 (0.817–0.978)		0.828 (0.561–0.939)	0.802 (0.507–0.929)
Inferior	0.856 (0.624–0.949)	0.839 (0.586–0.943)		0.891 (0.706–0.962)	0.870 (0.656–0.954)
Inferolateral	0.824 (0.553–0.937)	0.813 (0.530–0.933)		0.866 (0.648–0.953)	0.895 (0.717–0.964)
Anterolateral	0.860 (0.633–0.951)	0.815 (0.533–0.934)		0.843 (0.596–0.945)	0.829 (0.564–0.939)
Mid
Anterior	0.907 (0.746–0.968)	0.903 (0.735–0.966)		0.870 (0.657–0.954)	0.857 (0.626–0.949)
Anteroseptal	0.913 (0.761–0.970)	0.807 (0.517–0.931)		0.912 (0.759–0.970)	0.908 (0.748–0.968)
Inferoseptal	0.796 (0.495–0.927)	0.830 (0.567–0.940)		0.852 (0.616–0.948)	0.828 (0.561–0.939)
Inferior	0.839 (0.586–0.943)	0.810 (0.523–0.932)		0.850 (0.611–0.947)	0.886 (0.695–0.960)
Inferolateral	0.827 (0.560–0.938)	0.840 (0.589–0.943)		0.873 (0.664–0.995)	0.812 (0.528–0.933)
Anterolateral	0.874 (0.665–0.956)	0.785 (0.473–0.922)		0.837 (0.581–0.942)	0.857 (0.626–0.949)
Apical
Anterior	0.915 (0.766–0.971)	0.776 (0.453–0.919)		0.898 (0.725–0.965)	0.913 (0.762–0.970)
Septal	0.814 (0.531–0.933)	0.783 (0.467–0.921)		0.894 (0.714–0.963)	0.878 (0.676–0.957)
Inferior	0.830 (0.56–0.939)	0.841 (0.591–0.944)		0.840 (0.588–0.943)	0.768 (0.440–0.916)
Lateral	0.877 (0.673–0.957)	0.853 (0.617–0.948)		0.859 (0.630–0.950)	0.847 (0.604–0.946)

ECG, electrocardiogram; ICCs, intraclass correlation coefficients.

Discussion

This prospective method-comparison study systematically evaluated the performance of pulse-triggering versus ECG-triggering T1-mapping protocols at 3.0T cardiac MR. Our study revealed three main findings: (I) pulse triggering T1 mapping was feasible; (II) the image quality between pulse- and ECG-triggering protocols was equivalent (Likert score 4, indicating “good” diagnostic quality); (III) there were no significant differences in myocardial native T1 quantification between the two triggering modalities.

Cardiac MR is recognized as the gold standard non-invasive imaging tool in cardiovascular disease and continues to expand its utility in clinical practice (1). Conventional cardiac MR acquisitions rely on cardiac gating to capture information at end-diastole or end-systole (7). However, as clinical MRI systems increasingly adopt higher field strengths, elevated electromagnetic interference susceptibility and amplified MHD effects impair ECG signal fidelity, creating an urgent demand for alternative cardiac triggering schemes (24). As previous studies have shown, many methods could replace ECG triggering, such as pulse triggering (20), Doppler ultrasound triggering (25), and acoustic triggering (26).

Peripheral pulse triggering offers a cost-effective solution through PPG. Both the rise and fall of the PPG signal are associated with cardiac contraction and cardiac relaxation, making it suitable for cardiac triggering (19). Unlike ECG signals that suffer from T-wave amplification artifacts due to MHD effects at 3T and beyond, pulse-triggering acquisition eliminates electromagnetic interference risks. This fundamental physical distinction ensures reliable cardiac cycle detection regardless of the static magnetic field strength. In addition, it requires only a fingertip sensor, which simplifies the procedure, and is especially suitable for children or agitated patients. Previous studies have validated the efficacy of pulse triggering for LV functional assessment at 3T cardiac MR (20). We compared images acquired using finger-pulse triggering with those obtained using ECG triggering, and found that finger-pulse triggering can be used as an alternative and effective triggering method in T1 mapping.

The PPG waveform originates from cardiac mechanical activity, where LV contraction propels blood flow through the arterial system, generating peripheral pulse waves detectable via optical sensing. The PPG waveform lags behind the R wave in ECG (27). The time difference between the R peak in the ECG signal and the corresponding peak in the peripheral signal is called the PTT (28). In clinical practice, after recalculating and dynamically adjusting the TR acquisition period, pulse-triggered T1 mapping is still acquired at end-diastole. The present results show that the errors are negligible when using the PPG method during resting conditions (18,27). Compared with ultrasound triggering and acoustic triggering, pulse triggering is simpler to operate and cheaper, and the R-wave delay time is shorter. Lang et al. identified pulse-triggered signals as an alternative to ECG-triggered signals in gated myocardial perfusion single-photon emission computed tomography imaging, noting that the differences between the methods were not clinically significant (29). The results of our analysis showed that there was no significant difference in the quantitative analysis of T1 mapping obtained by pulse triggering and ECG triggering.

Admittedly, pulse triggering shares an apparent common drawback of conventional ECG, ultrasound, and acoustic triggering (i.e., extra hardware is required for signal detection and processing). To overcome the constraint of using ancillary hardware, various self-gating (SG) methods have been proposed and shown to be feasible in cardiac MR (30-32), especially for cardiac arrhythmias and children’s cardiac imaging. SG is a technique in which changes in the raw MR data over time are used to form a signal that may be used to retrospectively synchronize data to the cardiac cycle. The SG signal is typically produced by sampling the center of k-space repeatedly during cine data acquisition by interleaving the collection of non-phase encoded k-space lines (k_pe =0) (31). SG methods might be the most promising approach, especially for high-field magnets. However, due to its high algorithm complexity and low cost-effectiveness ratio, the popularization rate of SG in clinical practice is still limited.

Recent advances in ECG-free cardiac MRI techniques have enabled free-breathing myocardial tissue characterization. One study developed a free-breathing myocardial T1-mapping technique using the inversion-recovery radial fast low-angle shot and calibration-free motion-resolved model-based reconstruction, which had good accuracy and reproducibility (33). Another study introduced a free-running T1* mapping sequence that enables the non-ECG quantification of myocardial T1* changes during physiological exercise (34). Emerging free-running approaches now allow simultaneous T1/T2 mapping and cine imaging during free breathing, incorporating methods such as non-rigid cardiac motion correction for three-dimensional whole-heart imaging (35) and motion-corrected MR fingerprinting (36). These techniques eliminate the need for external synchronization (e.g., ECG gating), while maintaining diagnostic image quality during unconstrained respiratory and physiological motion conditions. Although these free-running technologies hold great promise, they remain limited by technical and computational challenges, and are still in the exploratory stages. This underscores the value of pulse triggering as a practical and interim solution.

In addition, in the presence of arrhythmias and heart rate variability, T1 mapping image quality is commonly sub-optimal and cannot meet clinical diagnosis requirements (37). Heart rate significantly affects MOLLI acquisition, particularly at lower rates (37,38). Compared with 3T or conventional MOLLI, 1.5T and MOLLI 5 s(3 s)3 s may better mitigate heart rate, respiratory, or cardiac motion influences (38,39). The saturation pulse-prepared heart rate-independent inversion-recovery sequence has been proposed as a heart rate-independent alternative, offering greater stability and shorter breath-hold requirements compared to conventional MOLLI sequences (37). T1 maps are obtained by acquiring multiple sample points on a longitudinal magnetization recovery curve after magnetization preparation (5). In bradycardia patients, the R-R interval is prolonged and the distribution of sampling points is sparse, which affects the accuracy of the exponential fitting and may increase the error of T1 mapping (38). Moreover, as the total sampling time increases, the difficulty of breath-holding may introduce respiratory artifacts. Hence, performing T1 mapping during systolic phase has recently been proposed in tachyarrhythmias (40), and has also demonstrated improved quality in patients with atrial fibrillation (41) and bradycardia (38). This study provided preliminary verification of the feasibility and stability of pulse triggering in cardiac MR T1 mapping. In the future, we will collect more data to explore the potential application of pulse-triggering T1 mapping in arrhythmias.

Limitations

The study had several limitations. First, this study was conducted at a single center, on a single magnet, from a single supplier, at a single field strength, and with a small sample size. Thus, the generalizability of the conclusions cannot be guaranteed, and future studies involving magnets of various field strengths from different suppliers and larger populations need to be conducted at multiple centers. Second, hematocrit values were not available for some patients in this study, so only native T1 values of myocardial tissue were validated (post-contrast T1 values were not validated), and no comprehensive assessment of extracellular volume (ECV) was conducted. Future studies will continue to determine whether pulse triggering can be used for the assessment of ECV. Third, our cohort excluded patients with severe peripheral vascular diseases or arrhythmias. We preliminarily validated the feasibility and stability of pulse triggering in T1-mapping scanning. Further studies need to be performed to evaluate whether pulse triggering is also robust in patients with arrhythmias and peripheral vascular diseases.

Conclusions

This prospective comparative investigation evaluated the interchangeability of peripheral pulse triggering versus conventional ECG triggering for T1-mapping acquisition on 3T cardiac MR, showing comparable T1 quantification accuracy. These findings support the use of pulse-triggering T1 mapping as a viable alternative and practical strategy for patients when ECG triggering is unfeasible. This method may also aid in follow-up and clinical interventions for patients in the future.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Tongji Hospital (IRB No. TJ-IRB20230914). Written informed consent was obtained from all participants prior to enrollment.

Data Sharing Statement

Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-900/dss

qims-15-10-9325-dss.pdf ^{(73.3KB, pdf)}

DOI: 10.21037/qims-2025-900