Free-breathing Cardiorespiratory Synchronized Cine MRI for Assessment of Left and Right Ventricular Volume and Function in Sedated Children and Adolescents with Impaired Breath-holding Capacity

Abstract

Purpose

To prospectively compare left ventricular and right ventricular volume, function, and image quality of a free-breathing (FB) cardiorespiratory synchronized balanced steady-state free precession cine MRI sequence with that of a standard of reference breath-hold (BH) technique in sedated children and adolescents who are unable to perform BHs.

Materials and Methods

Cohort 1 included 30 patients able to perform BHs (mean age, 19 years; age range, 9–69 years). Cohort 1 underwent both BH and FB cine short-axis imaging with identical acquisition parameters. Cohort 2 included 63 patients unable to perform BHs (50 sedated patients [mean age, 9 years; age range, 4 months to 28 years], 13 unsedated patients [mean age, 21 years; age range, 8–58 years]). Cohort 2 underwent FB cine imaging in multiple views with spatiotemporal resolution equivalent to BH imaging. Comparative quantitative analysis was performed for left ventricular and right ventricular volumes in cohort 1 and for qualitative image quality scores in all patients.

Results

Global left ventricular and right ventricular volumetric indexes and image quality scores were comparable between BH and FB sequences in cohort 1. FB image quality was graded as excellent (37 sequences), good (197 sequences), adequate (26 sequences), and suboptimal (three sequences) for 263 cine sequences in cohort 2. In cohort 1, de facto image acquisition time for FB (6.1 minutes ± 1.9 [standard deviation]) was comparable to the equivalent for BH (6.1 minutes ± 2.6) for a stack of 14 sections.

Conclusion

In cohorts of sedated children, adolescents, and young adults unable to perform BHs consistently, left ventricular and right ventricular volumes and function were comparable and image quality was noninferior between FB and standard of reference BH techniques.

© RSNA, 2019

Summary

A free-breathing cardiorespiratory synchronized cine sequence acquires the complete cardiac cycle with adequate spatial, contrast, and temporal resolutions for evaluating ventricular morphology, volume, and function with image quality equivalent to that of the standard of reference breath-hold sequence.

Key Points

• ■ Global left ventricular and right ventricular volumetric indexes and image quality scores were comparable between breath-hold (BH) and free-breathing (FB) sequences in cohort 1.

• ■ FB image quality was graded as excellent (14%), good (75%), adequate (10%), and suboptimal (1%) for 263 cine sequences in cohort 2.

• ■ In cohort 1, de facto image acquisition time for FB acquisition (6.1 minutes ± 1.9) was comparable to that of the equivalent BH acquisition (6.1 minutes ± 2.6) for a stack of 14 sections.

Introduction

Cardiovascular MRI is the current clinical reference standard for measurement of ventricular function and blood flow, which are both crucial components in the assessment of congenital heart disease (1). Routine clinical cine cardiovascular MRI requires suspension of respiration to obtain diagnostic image quality with adequate spatiotemporal resolution. This renders cardiovascular MRI–based assessment of complex, dynamic anatomy of smaller structures with faster heart and respiratory rates in children, adolescents, and young adults with congenital heart disease especially challenging. Specifically, within the pediatric population with cardiovascular disease, there are several factors that may limit the ability to perform consistent breath holds (BHs), such as anxiety, medical condition, and age. Young patients often require sedation or general anesthesia for cardiovascular MRI study because of their inability to lie still and perform consistent BHs. Hence, it is of significant clinical interest to safely obtain cine cardiovascular MR images with adequate spatial, temporal, and contrast resolutions over the complete cardiac cycle without the constraint of respiratory suspension.

Multiple free-breathing (FB) cine cardiovascular MRI approaches using data undersampling (2,3), real-time acquisition (3,4), and respiratory motion compensation (5–7) previously reported in the literature have certain limitations in their adoption for routine clinical pediatric practice: (a) requirement of contrast material administration (8–11), (b) requirement of considerably longer reconstruction times and/or advanced hardware capabilities (2–4,9,10,12), and (c) prospective cardiac gating and/or partial acquisition of the cardiac cycle (2,5,8). Diagnostic image quality and accuracy of ventricular volumetric and functional indexes of a FB respiratory-triggered balanced steady-state free precession (bSSFP) sequence, without contrast material administration and with real-time reconstruction, has been reported in pediatric and adult populations (7,13,14). The cardiorespiratory synchronized cine bSSFP sequence extends beyond the respiratory-triggered technique by incorporating prospective rejection of the data acquired during inspiration, along with the clinically established prospective cardiac arrhythmia rejection and retrospective cardiac gating. An FB respiratory-triggered technique was shown to provide significantly better image quality when compared with a multiple signal averaging technique in a pediatric population (13). To obtain diagnostic image quality, in terms of signal-to-noise-ratio, blood-to-myocardial contrast, and reduced banding artifacts, the bSSFP sequence requires uninterrupted (15) high-flip-angle (16) radiofrequency (RF) pulses with the shortest possible repetition time (17). These requirements lead to a high RF energy deposition rate, which increases the specific absorption rate. In routine clinical practice, the combination of the maximum possible flip angle and the minimum possible repetition time for cine bSSFP acquisition is constrained by the regulatory safety limit of 2 W/kg (18). Respiratory synchronized FB cine acquisitions have a longer duration of continuous RF energy deposition. The cardiorespiratory synchronized cine bSSFP sequence has been reported to significantly lower the specific energy deposition by 40% compared to respiratory-gated FB acquisitions in adults (19) and by 50%–60% compared to four signal averaging FB cine bSSFP acquisitions in pediatric patients (20).

The specific aims of this prospective study were to compare left and right ventricular volume, function, and image quality of a FB cardiorespiratory synchronized cine bSSFP sequence with those of a conventional BH technique and further evaluate its specific energy deposition and imaging time duration in a cohort of sedated children and adolescents and young adults unable to perform BHs consistently.

Materials and Methods

Study Protocol

This prospective study was approved by our institutional review board for assessing ventricular function in FB pediatric patients with a waiver for informed consent and complied with the Health Insurance Portability and Accountability Act of 1996. The FB cine bSSFP sequence in its current form was implemented within our institution, and all the data and information were always under the control of our institution. In this single-center study, 93 patients (mean age, 14 years; age range, 4 months to 69 years) undergoing cardiovascular MRI for various clinical indications, including assessment of ventricular volumes and function, were enrolled. In cohort 1, 30 consecutive patients (mean age, 19 years; age range, 9–69 years) able to follow breath-holding instructions and suspend respiration consistently were enrolled to the comparative arm of FB, with a BH sequence in the short-axis view. In cohort 2, 63 consecutive patients (mean age, 12 years; age range, 4 months to 58 years) either sedated or unable to perform consistent BHs were recruited for cine imaging with the FB sequence for all the clinically required cardiac views. All patients underwent clinical examination and were included in the analysis. Indications for cardiovascular MRI in cohort 1 (30 patients) and 2 (63 patients) were: tetralogy of Fallot (11 in cohort 1, seven in cohort 2), aortopathies (five in cohort 1, seven in cohort 2), transposition of the great arteries (two in cohort 1, five in cohort 2), pulmonary valve disease (zero in cohort 1, five in cohort 2), bicuspid aortic valve (two in cohort 1, three in cohort 2), myocarditis (one in cohort 1, four in cohort 2), cardiomyopathies (one in cohort 1, three in cohort 2), atrial septal defect (zero in cohort 1, five in cohort 2), ventricular septal defect (zero in cohort 1, five in cohort 2), double outlet right ventricle (two in cohort 1, two in cohort 2), partial anomalous pulmonary venous connection (zero in cohort 1, four in cohort 2), left ventricular noncompaction (one in cohort 1, two in cohort 2), hypoplastic right ventricle (one in cohort 1, one in cohort 2), chest pain (one in cohort 1, one in cohort 2), Fontan procedure (two in cohort 1, zero in cohort 2), truncus arteriosus (zero in cohort 1, two in cohort 2), atrioventricular septal defect (one in cohort 1, zero in cohort 2), Duchenne muscular dystrophy (zero in cohort 1, one in cohort 2), persistent fifth arch (zero in cohort 1, one in cohort 2), heart failure (zero in cohort 1, one in cohort 2), pericarditis (zero in cohort 1, one in cohort 2), sickle cell disease (zero in cohort 1, one in cohort 2), Kawasaki disease (zero in cohort 1, one in cohort 2), and right atrial mass (zero in cohort 1, one in cohort 2).

Data Acquisition

All cardiovascular MRI examinations were performed with a 1.5-T clinical imager (Ingenia, Philips Healthcare, Best, the Netherlands) using a 28-element phased-array coil for signal reception and vector electrocardiographic cardiac gating, with respiratory bellows placed over the mediastinum. In all 93 study participants, a series of vector electrocardiographically gated cine images was acquired as per the clinical indication in up to six different orientations, namely vertical long axis (five sections), four chamber (five sections), short axis (12–14 sections), left ventricular outflow tract (three sections), right ventricular outflow tract (three sections), aortic root (three sections), and axial (10–15 sections). All cine imaging was performed prior to administration of contrast agent. The FB cardiorespiratory synchronized cine bSSFP sequence schematically depicted in Figure 1 allows two FB modes: fixed, which involves single R-R interval per respiration (described previously [13,14,21]) and adaptive, which involves multiple R-R intervals per respiration with prospective rejection for inspiration during acquisition. Cohort 1 patients underwent BH cine imaging for all the cardiac views and fixed-mode FB imaging only for the short-axis sections, immediately before or following the BH short-axis imaging, with identical imaging parameters and section prescription. No special instructions were given for breathing during the FB imaging. Cohort 2 patients were imaged with the FB cine imaging sequence in all the clinically required cardiac views. Five experienced cardiovascular MR technologists autonomously chose which FB mode was to be used, evaluating the length of the stable expiration period in terms of number of cardiac cycles for a particular patient. The imaging parameters for both BH and FB cine imaging were as follows: repetition time msec/echo time msec, 2.5–2.7/1.25–1.35; flip angle, 60°; acquired voxel size, 1.6–2 × 1.6–2 × 4–8 mm³; acceleration factor, 1.3–2; acquired temporal resolution, 30–45 msec. Typical BH time was 6–8 seconds per section. Actual acquisition times, RF transmission, and physiology events were retrieved from imager log files. Effective RF duty cycle, effective specific absorption rate, and actual specific energy deposition were calculated by using logged data. In patients imaged with the adaptive mode, recorded respiratory cycles and R-R intervals were used to compute imaging duration in case these patients had been imaged with the fixed mode instead, hereon referred to as cFix mode.

Figure 1:

MRI sequences for cardiac-gated balanced steady-state free precession cine imaging. Breath hold (BH): In a conventional BH sequence, imaging data acquired during the first R-R interval (black boxes) are discarded. All subsequent data acquired with cardiac gating (green boxes) are used for image formation. Free breathing (FB): In the FB cardiorespiratory synchronized sequence, radiofrequency (RF) excitation commences after the detection of a respiratory trigger (inspiration or expiration). The data acquired until the time interval of R-top occurrence after the respiration trigger (t, double-headed arrows) is longer than a predefined duration τ, or time to attain steady state, are discarded (black boxes). In the fixed mode, data acquired during a subsequent R-R interval (green boxes) are used for image formation and RF excitations cease. In the adaptive mode, cardiac-gated data acquisition continues until the detection of an inspiration trigger, when RF excitations cease. The data from the last R-R interval that happened during inspiration are discarded (orange boxes), and the rest of the data (green boxes) are used for the image formation. Electrical signals from the cardiac leads (top blue line) and respiratory bellows (bottom blue line) are shown in blue. Both BH and FB sequences use prospective rejection of cardiac arrhythmias and retrospective cardiac gating. Phase-encoding steps for the multiphase segmented k-space acquisition are changed only after successful acceptance of the data. Red dot = expiration trigger, blue dot = inspiration trigger, black rectangle = data during approach to steady state are discarded, green rectangle = data are accepted for further processing, orange rectangle = data during inspiration are rejected, t = time after respiratory trigger, τ = time to attain steady state (450 msec), dotted line = threshold for respiratory phase change.

Image Analysis

All the short-axis data were analyzed offline by a cardiovascular MRI postprocessing expert with more than a decade of experience on a dedicated postprocessing workstation (CVI42, Circle Cardiovascular Imaging, Calgary, Alberta, Canada) as part of the clinical routine. The quantitative assessment of the left ventricular and right ventricular volumetric indexes (end-diastolic volume, end-systolic volume, stroke volume, ejection fraction) was finalized by a cardiovascular MRI reader clinically in charge of the patient. It involved five cardiovascular MRI readers, three pediatric radiologists and two pediatric cardiologists with a minimum of 5 years of experience in cardiovascular MRI. One pediatric radiologist (S.J.) and one pediatric cardiologist (C.N.) blinded to the study design graded the image quality of the cine images in all the views from all the cases in cohort 2 and short-axis sections with BH and FB sequences from cohort 1. The image quality scores were based on three main criteria: blood-to-myocardial contrast, endocardial edge delineation, and presence of motion artifacts throughout the cardiac cycle. The details of the image quality criteria are described elsewhere (14). Each criterion was graded on a scale of 1 to 5, where 1 is nondiagnostic, 2 is suboptimal but still diagnostic for volumetric analysis, 3 is adequate, 4 is good, and 5 is excellent. All the sections were reviewed individually, and the image quality score of the lowest quality section in a particular view was assigned to the entire stack in that view in each scoring criterion. The combined image quality score was calculated as the equal-weight average of the three scores, which underscores the overall performance of the sequence. All the patient data were included in the data analysis.

Data Analysis

Descriptive statistics of each of the left ventricular and right ventricular volumetric indexes for patients in cohort 1 (BH and FB) and cohort 2 (FB) are reported as mean ± standard deviation. Bland-Altman analysis (22) and the two-sided paired t test were used to compare each of the parameters computed using the BH sequence with those computed using the FB sequence in cohort 1 patients. The Kruskal-Wallis test was performed to compare group differences of patient population in terms of age, heart rate, and respiratory rate and also to compare differences in image quality scores between scoring criteria, between observers, and between cohort 1 and cohort 2. Wilcoxon signed-rank tests were performed to compare image quality scores assigned to the BH and FB sequences of cohort 1 patients. For each of the three image quality scoring criteria considered in the study, the percentage of clinical subjects who received a range of image quality scores was plotted as a bar graph. In addition, a similar bar graph was constructed with the combined score, computed as the equal-weight average of the three scores, to underscore the overall performance of the sequence on the basis of all three criteria. Two-sided paired t tests were used to compare imaging duration of actual acquisitions using adaptive mode with corresponding computed imaging time if these patients had been imaged with fixed mode (cFix). A P value less than .05 indicates a significant difference.

Results

In all 30 cohort 1 patients, both the standard of reference BH and FB cine bSSFP sequences were performed successfully in the short-axis view. Cine bSSFP imaging using the FB sequence was performed successfully in all 63 cohort 2 patients for a total of 263 sequences (63 short axis, 63 vertical long axis, 63 four chamber, 25 left ventricular outflow tract, 30 right ventricular outflow tract, and 19 aortic root). The mean heart rate was 85 beats per minute (range, 48–110 beats per minute), and the mean respiratory rate was 19 breaths per minute (range, 12–30 breaths per minute). Table 1 summarizes further detailed characteristics of the patients in each category.

Table 1:

Characteristics of the Study Population

Note.—Data are mean ± standard deviation or number of patients. BSA = body surface area, HR = heart rate, RR = respiration rate.

In cohort 1, de facto image acquisition time for FB (6.1 minutes ± 1.9) was comparable to that of the standard of reference BH (6.1 minutes ± 2.6) cine bSSFP imaging for a short-axis stack of 14 sections. Table 2 provides descriptive statistics, and Figure 2 depicts Bland-Altman plots comparing left ventricular and right ventricular volumetric indexes using FB and BH sequences in cohort 1 patients. Image quality scores for all 93 patients were comparable between two readers across all criteria, and the mean of two observers’ scores was used for further analysis. Mean rank scores of presence of motion artifacts for both FB and BH were significantly (P <.01) lower than those of blood-to-myocardial contrast for the FB sequence; mean rank scores of all other groups were comparable. The image quality scores were comparable within each individual scoring criteria, and combined image quality scores between FB and BH sequences were comparable. In 30 cohort 1 patients, the combined clinical score was excellent (20 of 30) to good (nine of 30) to adequate (one of three) for BH and excellent (19 of 30) to good (11 of 30) for FB (Fig 3). The combined image quality score was equal in 14 of 30 patients and greater in FB than BH in 10 of 30 of the cases, with the difference between FB and BH being less than 1 in all the cases. Figure 4 shows representative BH and FB images from cohort 1 patients with corresponding combined image quality score.

Table 2:

Left and Right Ventricular Volumetric Indexes and Difference in their Values between Breath-hold and Free-breathing Sequences in Cohort 1 Patients

Note.—Unless otherwise indicated, data are means ± standard deviations. BH = breath hold, BSA = body surface area, EDV = end-diastolic volume, EF = ejection fraction, ESV = end-systolic volume, FB = free breathing, LV = left ventricle, RV = right ventricle, SV = stroke volume.

Figure 2:

Bland-Altman plots compare left ventricular (LV) and right ventricular (RV) volumetric indexes between breath hold (BH) and free breathing. BSA = body surface area, FB = free-breathing cardiorespiratory synchronized balanced steady-state free precession cine MRI sequences. Dotted red lines = limits of agreement.

Figure 3:

Bar-plot analysis of image quality scores depicts percentage of patients who had image quality scores of excellent, good, adequate, suboptimal, or poor. A, B, Bar-plot analysis for each grading criteria based on blood-to-myocardial contrast (BMC), endocardial edge definition (Edef), and presence of motion artifacts (Mart) throughout the cardiac cycle for, A, breath hold (BH) and free breathing (FB) in cohort 1 patients, and, B, the fixed (Fix) and adaptive (Adp) modes in cohort 2 patients. The combined image quality score is the equal-weight average of the three scores, which underscores the overall performance of the technique. C, Bar-plot analysis of combined image quality scores in patients grouped by sedated (Sed), unsedated (Unsed), fixed mode, adaptive mode, and for all patients. Comb = combined image quality score.

Figure 4:

Representative breath-hold (BH) and free-breathing (FB) cardiac balanced steady-state free precession short-axis images in cohort 1 patients with corresponding combined clinical scores. The combined image quality score is the equal-weight average of scores in three main criteria: blood-to-myocardial contrast, endocardial edge delineation (Edef), and presence of motion artifacts throughout the cardiac cycle. Each criterion was graded on a scale of 1 to 5, where 1 is nondiagnostic, 2 is suboptimal but still diagnostic for volumetric analysis, 3 is adequate, 4 is good, and 5 is excellent. The patient in (1) had dark papillary muscles and endocardial trabeculae clearly visible with crisp Edef on the bright backdrop of the blood pool with both BH and FB. For the patient in (2), there was slight degradation of Edef with FB. In (3), slight degradation of Edef with BH was a little worse in FB. For the patient in (4), there was motion blurring in BH visible in the septum, which was improved with FB with only slight degradation of Edef. In (5), there was substantial motion artifact with poor Edef with BH; motion artifacts were significantly alleviated with FB with slight degradation of Edef. FB = free-breathing cardiorespiratory synchronized balanced steady-state free precession cine MRI sequences.

In cohort 2, 50 patients (mean age, 9 years; age range, 4 months to 28 years) were sedated and 13 patients (mean age, 21 years; age range, 8–58 years) were either unable to follow breathing instructions or could not perform BHs consistently. Of 63, 41 (33 sedated) were imaged with the fixed mode and 22 (17 sedated) were imaged with the adaptive mode of the FB sequence. Heart rate and respiratory rate were comparable among all the groups, while age for sedated patients was significantly lower (P < .01) than that of the unsedated patients. De facto image acquisition times for the fixed mode (6.2 minutes ± 1.8) and adaptive mode (6.1 minutes ± 2.0) FB cine bSSFP imaging for a short-axis stack of 14 sections were comparable to the standard of reference BH (6.1 minutes ± 2.6) from cohort 1. Overall, the imaging duration with the adaptive mode FB sequence (26.1 seconds per section ± 8.8) was 18% (P < .002) shorter than with cFix (31.8 seconds per section ± 12.2). In 11 (50%) of 22 patients imaged with the adaptive mode, the heart rate–to–respiratory rate ratio was higher than 4.7 and imaging time reduction with the adaptive mode FB sequence (23.7 seconds per section ±9) compared with cFix (37.54 seconds per section ± 12.9) was 30% ± 32 (Fig 5). The effective mean rate of RF energy deposition for the FB sequence was lower by 46.4% in the fixed mode (0.93 kJ/kg/min ± 0.17) and 58.6% in the adaptive mode (0.73 kJ/kg/min ± 0.12) compared with regulatory safety limit of 2 kJ/kg/min (Fig 5). Because of decreased RF duty cycle of the FB sequence (Fig 1), specific energy deposition per section was reduced in the fixed mode (18.96 J/kg ± 5.16) by 49.36% and in the adaptive mode (24.13 J/kg ± 9.55) by 39.24% compared with an equivalent sequence with continuous RF excitations with specific absorption rate of 2 kJ/kg/min.

Figure 5:

Box plots of imaging duration, specific absorption rate (SAR), and specific energy deposition (SED) for actual fixed (Fix in blue) and adaptive (Adp in red) modes of cardiorespiratory synchronized acquisition and for corresponding calculated imaging time and SED if the fixed mode was used instead of adaptive mode (cFix in magenta) using logged cardiac and respiratory rates. One-to-one line plots (dotted black lines) for imaging duration with adaptive and cFix modes depict a 30% reduction in imaging time for 11 (50%) of 22 patients imaged with the adaptive instead of the fixed mode who had a heart rate–to–respiratory rate ratio higher than 4.7. SED was also calculated in case four signal averaging (4NSA) had been used instead of fixed (F_4NSA in green) and adaptive (A_4NSA in green) mode, depicting reduction in radiofrequency dose due to reduced duty cycle with cardiorespiratory sequence. Center line = median, whiskers = minimum and maximum within 1.5 times interquartile distance, * = outliers beyond 1.5 times the interquartile distance.

The combined clinical scores were excellent (37 [14%] of 263), good (197 [75%] of 263), adequate (26 [10%] of 263), and suboptimal (three [1%] of 263) in 263 FB cine sequences in cohort 2 (Fig 3). The image quality scores were comparable among fixed, adaptive, sedated, and unsedated groups. Image quality scores of the FB sequence for cohort 2 patients were significantly lower (P < .01) than those for cohort 1 patients for both FB and BH sequences. Figure 6 shows representative FB images from cohort 2 patients with excellent, good, and adequate combined image quality score.

Figure 6:

Representative free-breathing cardiac balanced steady-state free precession short-axis images in cohort 2 patients with combined clinical scores greater than 4.5 (excellent), between 4.5 and 3.5 (good), and between 3.5 and 2.5 (adequate). The combined image quality score is the equal-weight average of scores in three main criteria: blood-to-myocardial contrast (BMC), endocardial edge delineation (Edef), and presence of motion artifacts throughout the cardiac cycle. Each criterion was graded on a scale of 1 to 5, where 1 is nondiagnostic, 2 is suboptimal but still diagnostic for volumetric analysis, 3 is adequate, 4 is good, and 5 is excellent. The patient in (1) had dark papillary muscles and endocardial trabeculae clearly visible with crisp Edef on the bright backdrop of the blood pool. The patient in (2) had slight degradation of Edef in a couple of sections. For the patient in (3), there was slight degradation of Edef and slight blurring of trabeculae. For the patient in (4), there was slight degradation of Edef and slight reduction in BMC. In (5) and (6), there was slight motion blurring in a couple of sections along with slight degradation of Edef. For the patient in (7), there was substantial motion artifact with poor Edef.

Discussion

The primary finding of this prospective study was that FB cardiorespiratory synchronized cine MRI is feasible and produces left ventricular and right ventricular volumetric indexes comparable to those of standard of reference BH cine MRI. The FB cardiorespiratory synchronized cine bSSFP sequence provides image quality and spatiotemporal resolution equivalent to a BH sequence, which is key for assessment of wall motion abnormalities such as hypokinesis, akinesis, and dyskinesis, particularly in the context of smaller structures and rapid heart rates encountered in the pediatric population. Assessment of wall motion abnormalities is crucial in diagnoses such as arrhythmogenic right ventricular cardiomyopathy and myocarditis. The FB sequence performs prospective arrhythmia rejection and retrospective cardiac gating identical to the standard of reference BH sequence; this captures the active filling phase of the ventricles, which is crucial for accurate assessment of ventricular volumes as well as presence of diastolic dysfunction. Limiting the degree of blurring associated with multiple signal averaging can have critical importance in pathologic conditions such as hypertrophic cardiomyopathy where accurate wall thickness measurements are required and increased trabeculations of the left ventricle requiring accurate measurement of the compacted and noncompacted myocardium. Accurate estimation of volumetric and functional indexes is adversely affected by the nonphysiologic anesthetized state, as well as during BH sequences that alter intrathoracic pressure. Limiting the use of anesthetics has become of higher concern given recent studies that demonstrate adverse neurologic outcomes in very young children subjected to multiple anesthesia events (23). A cardiorespiratory synchronized cine bSSFP sequence allows cardiovascular MRI studies with lighter sedation, and this is further highlighted by the number of patients who completed the study without use of anesthetics. In addition, the FB sequence allows imaging to occur in a normal physiologic state, which can be critical in complex physiology such as patients who have undergone total cavopulmonary connection (24).

A secondary finding has the significant reduction in specific energy deposition with effective specific absorption rate less than 60% of the regulatory safety limit of 2 W/kg. This indicates benefits of interrupted RF excitations in the FB sequence that permits the shortest feasible repetition time at a high flip angle to achieve optimal blood-to-myocardial contrast with reduced banding artifacts. One of the considerations in cardiovascular MRI examination is the rise in core body temperature due to thermal dose incurred during the imaging session. Pediatric cardiovascular patients may have impaired thermoregulatory response compounded by small body size (25), young age (26), and poor cardiac output (27,28). The approach of multiple signal averaging to compensate for respiratory motion over multiple respiratory cycles leads to spatial blurring and is associated with higher thermal dose during the prolonged continuous data acquisition. Thus, it is prudent to pay special attention to specific absorption rate and specific energy deposition in pediatric patients, particularly those who have compromised cardiac output. Temperature instability may contribute negatively to cardiovascular status, particularly in those with underlying cardiac disease (29). Additionally, the adaptive mode allows data acquisition over consecutive cardiac cycles during expiration with prospective constraint for rejecting the data acquired during inspiration. Thus, it shortens the imaging duration in patients with higher heart rate–to–respiratory rate ratio, in which case the fixed mode would have resulted in prolonged imaging duration. Although there was no statistically significant difference in heart rate–to–respiratory rate ratio between patients imaged with the fixed mode and those imaged with the adaptive mode, the adaptive mode of the FB sequence reduced imaging duration by 18% (range, 0%–75%) compared with the fixed mode. In combination with the ratio of heart rate to respiratory rate, the pattern of breathing in terms of shallow versus sharp transitions governed the technologists’ choice of adaptive mode versus fixed mode. Limiting imaging time is critical in younger patients undergoing sedation and also having difficulty in thermoregulation. The ability to breathe freely and avoid the high thermal doses associated with multiple signal averaging is important.

Another advantage of the cardiorespiratory synchronized cine bSSFP sequence is that it could potentially reduce the errors in estimation of ventricular chamber volumes resulting from inconsistent BHs. Although cardiovascular MRI is considered the reference standard for the quantification of ventricular chamber volumes, inconsistencies in the level of the diaphragm across BHs for each section can result in under- and/or oversampling of the myocardium in the direction of the ventricular long axis. This misregistration may lead to variability in calculated volumes and ejection fraction, especially when involving sampling-dependent inclusion or exclusion of the basal section in end systole (30). This problem would be substantially diminished with FB acquisitions (5).

A potential limitation of the study was that cohort 2 patients were substantially younger than cohort 1 patients, primarily due to difficulty in obtaining patients with the ability to perform consistent BHs in that age group. Equivalence of volume and function proven in cohort 1 is suggestive evidence of accuracy of these volumetric parameters in a younger population. Another limitation of the study was that the core temperatures were not measured explicitly; however, actual RF deposition was recorded. One of the limitations of the FB cardiorespiratory synchronized cine bSSFP sequence is that it can get prolonged in the case of irregular heart rhythm which is an inherent limitation of cine imaging with prospective arrhythmia rejection and retrospective cardiac gating. Additionally, in cases of bradycardia whereby the heart rate is relatively low compared with the respiratory rate, data rejection rate of the adaptive mode is higher due to limited expiratory time limiting the attainment of the steady state and the acquisition of the complete cardiac cycle. In this scenario, the fixed mode with inspiration trigger and reduced preparation time shortens imaging duration, but may result in hyperintensity at the beginning of the cardiac cycle due to limited time to get to steady state and motion artifacts due to irregular breathing. The cardiorespiratory synchronization approach can be seamlessly combined with real-time cine techniques such as noncartesian sparse sampling to mitigate these limitations.

In conclusion, FB cardiorespiratory synchronized cine MRI produces accurate evaluation of ventricular morphology and volumetric indexes in sedated patients and patients with poor breath-holding ability, providing comparable image quality and spatiotemporal resolution if conventional BH imaging had been feasible. Image quality of FB sequence is noninferior to conventional BH imaging and superior to multiple signal averaging with reduced RF dose and added value of shorter sedation or no sedation.