Reducing respiratory artifacts in gadoxetic acid-enhanced magnetic resonance imaging via a patient-adapted breath-holding training strategy

Introduction

Gadoxetic acid is a widely used liver magnetic resonance (MR) contrast agent with the ability to diagnose hepatocellular carcinoma (1,2), metastatic liver cancer (3-5), and other diseases effectively and specifically (6) due to its selective uptake and extracellular distribution during the hepatobiliary phase (7-9). The degree of hepatobiliary enhancement can serve as an indicator of liver function (10). Despite the broad utility of gadoxetic acid in diagnosing liver disease, arterial phase (AP) artifacts in gadoxetic acid-enhanced imaging represent a significant weakness that are lacking in other extracellular gadolinium-based contrast agents (11,12). In gadoxetic acid imaging, especially in individuals with chronic liver disease, the hepatic AP is crucial for characterizing localized liver lesions. However, the image quality is severely degraded due to transient severe motion (TSM) and truncation artifacts (13-16). TSM tends to occur during the late AP, likely due to a sharp increase in the peak plasma concentration of gadoxetic acid, which is associated with acute dyspnea and compromises the diagnostic evaluation of focal liver lesions (16).

Thus far, various strategies have been proposed for overcoming AP artifacts in gadoxetic acid imaging. First, low-flow rate injection protocols have been used to avoid the ringing and ghost artifacts related to rapidly increasing signal intensity (SI) in the center of k-space during a short signal acquisition interval (17,18). Second, fast or sparse acquisition and deep learning-based image filter techniques have been used to reduce respiratory motion artifacts (19-21). Third, 50%-gadoxetic acid dilution has been used to reduce the dyspnea caused by dose toxicity (16,22,23). Finally, a spiral dilution strategy consisting of a spiral flow-generating tube and modified breathing commands, which are more patient-friendly, has been applied to improve the quality of AP imaging (24,25).

The improvement in MR and high-pressure injector technologies may be able to effectively reduce the artifacts of AP images. However, these aforementioned artifact reduction strategies rely more heavily on advancements in MR imaging (MRI) hardware, sequence development, and high-pressure injector technology. These areas remain subjects of ongoing research and have not yet been widely adopted in clinical institutions, particularly in primary care hospitals. At present, the widely accepted form of injection methods for gadoxetic acid imaging are a slower injection protocol at an injection rate of 1 mL/s without saline dilution or a 50% contrast agent dilution at the injection rate of 2 mL/s (16,22,23,26). Research in recent years has indicated that a 1:1 saline dilution of gadoxetic acid combined with a slow injection rate of 1 mL/s can significantly reduce AP artifacts for patients undergoing at least two serial gadoxetic acid-enhanced liver MRI scans (16,27). This new finding suggests that two of the most promising injection protocols—one using a 1:1 dilution of gadoxetic acid with normal saline and an injection rate of 2 mL/s and the other using undiluted gadoxetic acid injected at 1 mL/s—may be further optimized. Thus, on the basis of the suggested gadoxetic acid-based AP image injection protocols, we attempted to develop a patient-adapted breath-holding training strategy that may be more suited to promoting and reducing respiratory artifacts for patients’ first gadoxetic acid examination.

Previous reports concerning the reduction of AP artifacts during gadoxetic acid imaging have not examined patient breath-training strategies, and there is insufficient evidence to prove whether it is effective in reducing TSM. To our knowledge, the majority of gadoxetic acid imaging methods require patients to perform a breath-hold, but no studies have examined reducing patient fatigue to mitigate TSM. Consequently, this study assessed the feasibility of combining a modified breath-training method and injection protocols to obtain high-quality AP images. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1225/rc).

Methods

Patients

This single-center retrospective study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by the Ethics Board of The First Affiliated Hospital of Army Medical University, Chongqing, China [No. (B)KY2022122]. The requirement for informed consent was waived due to the retrospective nature of the analysis. We reviewed the picture archiving and communication system (PACS) and records for breath-holding training strategy for patients who underwent gadoxetic acid-based upper abdomen examination via a 3.0T MR system in The First Affiliated Hospital of Army Medical University between May 2019 and February 2020. Based on our data collection and survey results regarding the breath-holding training habits of nurses (prior to gadoxetic acid-enhanced MR examination, each patient is required to undergo breath-holding training by the nurse on duty in our daily work), we found that two models are primarily used: one increases the breath-holding time from 5 seconds to the sequence scan time gradually (gradual breath-holding training strategy) and the other matches the breath-holding time to the sequence scan time (nongradual breath-holding training strategy). Over the examined study period, patients were instructed to hold their breath after deep inspirations, as most of the nurses agreed that this approach has been more effective in the past few years. Due to the limitations of this study’s retrospective design, we excluded unrecorded or ambiguous data from 1,387 gadoxetic acid examinations. The inclusion criteria for patients were as follows: (I) no previous experience related to undergoing gadoxetic acid-enhanced upper abdomen MRI and (II) no discomfort or adverse symptoms after gadoxetic acid injection. The exclusion criteria for this study are summarized in Figure 1. According to the inclusion and exclusion criteria, 126 patients were ultimately included in the study (Figure 1). Over the study period, our institution implemented a shift system. The evaluations in this study were based on the injection protocols used by radiographers and the respiratory training strategies employed by nurses. Throughout the 10-month period, only four radiographers consistently adhered to internationally recognized injection protocols, while two respiratory training nurses maintained consistent practices that could be reviewed.

Figure 1 Flowchart of patient inclusion in this study. MRI, magnetic resonance imaging.

The 126 patients enrolled in this study were divided into four groups according to whether the gadoxetic acid was diluted and the type of breath-holding training strategy. In group 1, 31 patients underwent training with gradual increases in breath-holding time from 5 seconds to 20 seconds, and 50% diluted gadoxetic acid was administered at a rate of 2 mL/s. In group 2, 32 patients underwent training with a nongradual breath-holding time of 20 seconds under the same injection strategy as that of group 1. In group 3, 30 patients underwent the same breath-holding training protocol as that in group 1 but with undiluted gadoxetic acid at an injection rate of 1 mL/s. For group 4, 33 patients underwent the same breath-holding training protocol as that in group 2 with the same injection strategy as that in group 3.

All patients underwent a 20-mL saline flush, with the same gadoxetic acid injection rate being applied in each group. The patients’ basic characteristics are summarized in Table 1. Figure 2 displays examples of the respiratory artifacts on AP images in each group.

Figure 2 Examples of respiratory artifacts on arterial-phase images in each group. (A) A 57-year-old man with hepatitis from group 1; respiratory artifact score =3; hepatic artery SNR =1.58; portal vein SNR =1.34; hepatic parenchyma SNR =0.97; CNR for the hepatic artery to liver parenchyma =0.62; CNR for the hepatic artery to portal vein =0.25; CNR for the portal vein to liver parenchyma =0.37. (B) An 82-year-old man with hepatic carcinoma from group 2 with best image quality: respiratory artifacts score =1; hepatic artery SNR =3.21; portal vein SNR =2.64; hepatic parenchyma SNR =1.51; CNR for the hepatic artery to liver parenchyma =1.69; CNR for the hepatic artery to portal vein =0.57; CNR for the portal vein liver parenchyma =1.13. (C) A 49-year-old man with hepatitis from group 3 with a severe respiratory artifact; respiratory artifact score =3; hepatic artery SNR =1.40; portal vein SNR =2.15; hepatic parenchymal SNR =0.96; CNR for the hepatic artery to liver parenchyma =0.37; CNR for the hepatic artery to portal vein =0.75; CNR for the portal vein to liver parenchyma =1.12. (D) A 47-year-old man with hepatitis from group 4 without obvious artifact; respiratory artifact score =2; hepatic artery SNR =1.11; portal vein SNR =1.48; hepatic parenchymal SNR =1.33; CNR for the hepatic artery to liver parenchyma =0.22; CNR for the hepatic artery to portal vein =0.38; CNR for the portal vein to liver parenchyma =0.15. The artifact was severe in (A) and (C) but not in (B) and (D). However, the SNR of portal vein and the CNR of portal vein to liver parenchyma was the best in (B). Group 1, gradual strategy + diluted injection; group 2, nongradual strategy + diluted injection; group 3, gradual strategy + undiluted injection; group 4, nongradual strategy + undiluted injection. CNR, contrast-to-noise ratio; SNR, signal-to-noise ratio.

Breath-hold training protocol

The protocol for the gradual breath-holding training strategy consisted of four sequential stages requiring patients to breath-hold for 5, 10, 15, and 20 seconds, respectively. If a patient failed to meet the criterion at any stage, the stage was repeated until successful completion.

The protocol for nongradual breath-holding training strategy required patients to breath-hold for 20 seconds (equivalent to the fourth stage of the gradual protocol). Training was repeated until the criterion was met.

A successful breath-hold was defined as the complete absence of abdominal motion as assessed through manual palpation and visual observation, with the examiner’s hand placed on the patient’s abdomen to monitor movement during breath-holding.

The rest periods between tasks were determined subjectively by the patients. The following task began only when the patient felt sufficiently recovered to proceed.

MR examination protocol

Imaging of the abdomen was performed with a MAGNETOM Trio 3T MRI scanner (Siemens Healthineers, Erlangen, Germany) with a six-channel body coil. Prior to the examination, an 18-gauge disposable medical indwelling needle was inserted into the patient’s peripheral vein and was followed by breath-holding training. During the examination, the patients were placed in the supine position, and the MRI procedure was performed under fasting conditions. The indwelling needle was connected to a high-pressure injector (MEDRAD Spectris Solaris EP, Bayer, Leverkusen, Germany). A standard dose (0.025 mmol/kg body weight) of gadoxetic acid (Primovist, Bayer) was administered and followed by 20-mL saline flush. The conventional imaging protocol prior to dynamic enhancement was applied and included a coronal half-Fourier acquisition single-shot turbo spin-echo (HASTE) sequence (acquired with two breath-holds), an axial HASTE sequence (also acquired with two breath-holds), and an axial two-dimensional fast low-angle shot (FLASH) T1-weighted in-phase and out-of-phase sequence (acquired in a single breath-hold). The acquisition time for hepatic AP images was confirmed via MR fluoroscopic triggering technology. A scan was initiated when the SI of the upper abdominal aorta showed obvious enhancement. Pre- and postcontrast three-dimensional volumetric interpolated breath-hold examination (3D-VIBE) sequences were used. After injection, contrast-enhanced MRI was initiated at 60 s for the portal venous phase; at 180 s for the equilibrium phase; and at 5, 10, and 15 min for the multiphasic hepatocyte phases. Table 2 displays each sequence’s parameters.

Image analysis

All of the AP images were evaluated with double-blind scoring and were analyzed by two radiologists (Yuting Wang, MBBS and Jian Wang, PhD, with 20 and 36 years of abdominal diagnostic experience, respectively) using a 5-point grading approach. The two radiologists evaluated the AP image quality as follows: (I) no artifact; (II) minimal artifacts without effects on diagnosis; (III) moderate artifacts with limited effects on diagnosis; (IV) severe artifacts in images that remained interpretable; and (V) extensive artifacts that prevented diagnosis (22).

The average scores across two radiologists were used for statistical analysis.

For quantitative analysis, one radiologist drew regions of interest (ROIs) on the axial VIBE sequence in the AP position to measure the SIs of the following structures: the hepatic artery, portal vein, liver parenchyma, and paraspinal muscle. The selection methods of ROI were based on previous research (16). One radiologist with more than 5 years of experience in abdominal MR diagnosis and blinded to the injection methods and breath-holding training strategies conducted the measurement of SI areas by placing the ROIs at the level of celiac trunk and the levels above and below celiac trunk, respectively. The mean SI of liver parenchyma was measured in the right, left medial, and left lateral lobes, with liver lesions and vessels with a circular area larger than 2 cm² being avoided. The SIs of the portal vein and hepatic artery were placed in the main branches based on the shape of the blood vessels. The mean and standard deviation (SD) of the paraspinal muscle were defined as the areas at the paraspinal muscles on the left and right sides delineated with a circular ROI of about 12 cm². All ROIs were measured three times on three consecutive planes, with the largest acceptable ROI size being used. To obtain the highest possible SI, the SI of all tissues was assessed with a circular cursor in the increased region, with cystic or necrotic areas being excluded.

The signal-to-noise ratio (SNR) of the portal vein, hepatic artery, and liver parenchyma were calculated by dividing the SI of each organ by one SD of the paraspinal muscle SI as follows: SNR = SI of the organ/SD of paraspinal muscle SI.

The contrast-to-noise ratio (CNR) for the hepatic artery to portal vein, portal vein to liver parenchyma, and hepatic artery to liver parenchyma were calculated by dividing the difference between the SI of the organ A (liver parenchyma, hepatic artery, or portal vein) by that of the SI of the organ B (liver parenchyma, hepatic artery, or portal vein) with one SD of the paraspinal muscle SI. The difference between the SI of different the organ A and B were calculated as the absolute value as follows:

CNR = absolute value (SI of the organ A − SI of the organ B)/SD of paraspinal muscle SI (organ A ≠ organ B)

Statistical analysis

The Chi-squared test was used to calculate the distribution differences in sex distribution and disease types across the patients in the four groups. Differences in age, body weight, and body mass index were evaluated via one-way analysis of variance. The Kruskal-Wallis test was used for comparisons of multiple samples to identify differences in the abdominal aorta, hepatic artery, portal vein, hepatic parenchyma enhancement, respiratory artifacts, and total scores of the five components. Patients with liver cancer, hemangioma, hepatitis, and cirrhosis were tested with the corrected Chi-squared test. Differences in hepatic artery SNR, portal vein SNR, CNR of the liver parenchyma to hepatic artery, CNR of the portal vein to hepatic artery, and CNR of the portal vein to the liver parenchyma were calculated with the Kruskal-Wallis test. If the P value from the Kruskal-Wallis test indicated a significant difference (P<0.05), the pairwise comparisons between the two groups were confirmed with the Mann-Whitney test. The adjusted Chi-squared test and the Mann-Whitney test were conducted to assess the difference in TSM incidence and respiratory artifact scores between the gradual and nongradual groups, respectively. All statistical analyses were performed in R software version 4.1.3 (The R Foundation for Statistical Computing, Vienna, Austria). Data are expressed as the mean ± SD. The threshold for statistical significance was P=0.05.

Results

The four groups did not differ in terms of age (P=0.976), sex distribution (P=0.296), body weight (P=0.741), or body mass index (P=0.922) or in the proportions of liver cancer (P=0.532), hemangioma (P=0.381), hepatitis (P=0.468), and cirrhosis (P=0.845) (Table 1).

According to the semiquantitative evaluation analysis, the four groups also did not differ in terms of the scores for the hepatic artery (P=0.177), portal vein (P=0.214), abdominal aorta (P=0.599), and liver parenchyma (P=0.243) or the total score (P=0.235) (Table 3). However, respiratory artifacts significantly differed among the groups (P=0.002). The scores of artifacts were significantly lower in group 2 than in group 1 (1.53±0.62 vs. 2.56±1.16; P<0.001), lower in group 4 than in group 1 (1.79±0.89 vs. 2.56±1.16, P=0.006), lower in group 2 than in group 3 (1.53±0.62 vs. 2.12±1.08, P=0.037), similar between groups 1 and 3 (P=0.116), similar between groups 2 and 4 (P=0.327), and similar between groups 4 and 3 (P=0.236) (Table 3 and Figure 3). Group 2’s gadoxetic acid was diluted 50% with normal saline and administered at 2 mL/s according to the mean score of the blinded evaluation conducted by two physicians. The intervention comprising a nongradual breath-holding time of 20 s without abdominal movement produced the fewest transient respiratory artifacts in the upper abdominal AP while providing the best image quality. Patients in group 4 (undiluted gadoxetic acid at 1 mL/s and a nongradual breath-time of 20 s without abdominal movement) exhibited considerably fewer respiratory artifacts than did patients in group 1. Although there was no statistically significant difference between groups 2 and 4, group 2 had a slightly better respiratory artifact score. Additionally, the incidence of TSM was significantly higher in the gradual breath-holding group (groups 1 and 3; 8/61) compared to the nongradual group (groups 2 and 4; 1/65), as assessed by the adjusted Chi-squared test (P=0.030). The respiratory artifact score was significantly higher in patients in the gradual breath-hold training group (groups 1 and 3) than in those in the nongradual breath-hold training group (groups 2 and 4) (2.34±1.13 vs. 1.66±0.78; P<0.001).

Figure 3 Comparison of respiratory artifacts between each group. Artifacts were significantly smaller in group 2 than in group 1 (1.53±0.62 vs. 2.56±1.16; P<0.001), smaller in group 4 than in group 1 (1.79±0.89 vs. 2.56±1.16; P=0.006), and smaller in group 2 than in group 3 (1.53±0.62 vs. 2.12±1.08; P=0.037). Group 1, gradual strategy + diluted injection; group 2, nongradual strategy + diluted injection; group 3, gradual strategy + undiluted injection; group 4, nongradual strategy + undiluted injection.

The four groups did not significantly differ in the quantitative scores for the SNR of the hepatic artery (P=0.909), the CNR of the hepatic artery to liver parenchyma (P=0.980), and the CNR of the hepatic artery to portal vein (P=0.532). The SNR of the portal vein differed significantly between groups 1 to 4 (1.95±0.50, 2.07±0.53, 1.59±0.48, and 1.75±0.50, respectively; P<0.001): group 2 had a significantly greater SNR than did group 3 (2.07±0.53 vs.1.59±0.48; P<0.001) and group 4 (2.07±0.53 vs. 1.75±0.50; P=0.007). Additionally, the portal vein SNR of group 1 was significantly higher than that of group 3 (1.95±0.50 vs. 1.59±0.48; P=0.004) (Figure 4). The analysis of the SNR for the liver parenchyma was conducted via the Mann-Whitney test. Group 2 had a significantly higher SNR than did group 3 (1.28±0.26 vs. 1.14±0.17; P=0.004) (Figure 5). The CNR of the portal vein to liver parenchyma differed significantly across groups 1 to 4 (0.73±0.39, 0.79±0.41, 0.51±0.38, and 0.58±0.29, respectively; P=0.008), and that in group 2 was higher than that in group 3 (0.79±0.41 vs. 0.51±0.38; P=0.004) and group 4 (0.79±0.41 vs. 0.58±0.29; P=0.020) and higher in group 1 than in group 3 (0.73±0.39 vs. 0.51±0.38; P=0.021) (Table 4 and Figure 6).

Figure 4 Comparison of the SNR of the portal vein between the groups. The SNR of the portal vein was confirmed via the Mann-Whitney test. The SNR in group 2 was significantly higher than that of group 3 (2.07±0.53 vs. 1.59±0.48; P<0.001) and group 4 (2.07±0.53 vs. 1.75±0.50; P=0.007). The SNR of group 1 was significantly higher than that of group 3 (1.95±0.50 vs. 1.59±0.48; P=0.004). Group 1, gradual strategy + diluted injection; group 2, nongradual strategy + diluted injection; group 3, gradual strategy + undiluted injection; group 4, nongradual strategy + undiluted injection. SNR, signal-to-noise ratio.

Figure 5 Comparison of the SNR of the liver parenchyma between the groups. The SNR of liver parenchyma was confirmed by the Mann-Whitney test. The SNR of group 2 was significantly higher than that of group 3 (1.28±0.26 vs. 1.14±0.17; P=0.004). Group 1, gradual strategy + diluted injection; group 2, nongradual strategy + diluted injection; group 3, gradual strategy + undiluted injection; group 4, nongradual strategy + undiluted injection. SNR, signal-to-noise ratio.

Figure 6 Comparison of the CNR of the portal vein to the liver parenchyma between the groups. The CNR of the portal vein to liver parenchyma was significantly higher in group 2 than in group 3 (0.79±0.41 vs. 0.51±0.38; P=0.004) and group 4 (0.79±0.41 vs. 0.58±0.29; P=0.020) and higher in group 1 than in group 3 (0.73±0.39 vs. 0.51±0.38; P=0.021). Group 1, gradual strategy + diluted injection; group 2, nongradual strategy + diluted injection; group 3, gradual strategy + undiluted injection; group 4, nongradual strategy + undiluted injection. CNR, contrast-to-noise ratio.

Quantitative scoring analysis revealed that the patients in group 2 (50% gadoxetic acid diluted with normal saline and injected at 2 mL/s), who used a nongradual breath-holding training strategy of direct breath-holding time for 20 s without abdominal undulation, had the highest portal vein SNR and CNR of the portal vein to liver parenchyma in the upper abdominal AP, which was more conducive to diagnosis.

Discussion

These findings of this study indicate that the breath-hold training strategy, as part of the MRI protocol, is a potential factor that can impact image quality in the AP phase of gadoxetic acid-enhanced liver MRI. The image quality of the gadoxetic acid-enhanced hepatic AP was slightly improved by contrast agent dilution as reflected in the lower respiratory artifacts scores of group 2 compared to group 4, which is consistent with previous studies (22,23,26). A transiently high non-diluent gadoxetic acid level in serum has been speculated to result in hyperventilation due to the triggering of central chemoreceptors (23). As image quality in patients with both gradual and nongradual training strategies was significantly improved, we believe that injection of gadoxetic acid diluted by 50% is feasible and beneficial for MR examination.

Respiratory artifacts were reduced in patients subjected to the nongradual breath-holding training method: the score was lower in group 2 than in group 1 and was slightly lower in group 4 than in group 3. Moreover, the incidence of TSM and respiratory artifacts was significantly higher in the gradual breath-holding group (groups 1 and 3) than in the nongradual breath-holding group (groups 2 and 4). These results suggest that the nongradual breath-holding strategy may not only reduce respiratory artifacts in gadoxetic acid-enhanced MRI but also lower the incidence of TSM, making it potentially more suitable for patients. It is possible that the gradual breath-holding training method is more energy-demanding for patients, causing physical distress and boredom. Therefore, this study suggests that breath-holding training strategy can play an important role in the severity of respiratory artifacts. For two of the most recommended gadoxetic acid AP acquisition methods (16), a patient-adapted breath-holding training strategy may be more effective than optimizing concentration and dose to in reducing respiratory artifacts. Quantitative analysis revealed that the nongradual breath-training method improved the SNR of the portal vein and the CNR of the liver parenchyma to the portal vein; these values were optimal in group 2. The nongradual breath-holding training method, as compared with the gradual breath-holding training method, was more appropriate for AP gadoxetic acid MRI. It may be more complicated for the patients to use breath-holding training method, with breath-holding exercises repeated from 5 to 20 s. Westbrook et al. showed that an individual’s task performance can be affected by distractions and multitasking (28). Nongradual breath training is the final task in the gradual breath-training method. Gradual breathing training method involves three more tasks than does nonprogressive breathing training. Therefore, we believe that gradual breathing training method requires greater physical exertion and is more likely to lead to accumulated fatigue. Although we provide sufficient rest time for patients during gradual breath-hold training, the psychological stress on patients may still increase due to the large number of tasks and repetitive exercises, potentially leading to boredom and distraction. According to literature, there is an optimal respiratory rate and tidal volume at which the work of breathing is minimized (29). Gradual breath-hold training may deviate from this optimal range, resulting in a continuous increase in respiratory workload. This elevates the pressure generated by the inspiratory muscles and raises the energy demand, leading to the accumulation of inspiratory muscle fatigue. Furthermore, lactic acid deposition in the muscles is cleared slowly, delaying recovery. These factors may explain why gradual breath-holding strategies produce higher levels of respiratory artifacts and TSM and consequently poorer image quality as compared to nongradual breath-holding strategies.

Another possible explanation for this observation involves the interaction between stress and recovery states. One review found that when pressure increases, the required recovery time also increases (30). Repeated training in the gradual breath-training group increased patient stress and did not provide sufficient physical recovery time. Such overtraining can lead to burnout and depression, reducing the value of training (31,32). In the nongradual breath-training group, patients entered a state of meditative rest after breath-holding. Meditation and relaxation occur simultaneously. Because of sympathetic nerve activity, the response ability to stimuli after relaxation provided a significant temporary improvement for the completion of the breath-holding task (33,34). Furthermore, each patient must be able to complete the nongradual task prior to undergoing the MRI examination. All patients are required to perform two breath-hold acquisitions during the noncontrast scan, which is equivalent to a repetition of the no-gradual task. Since this procedure is free from the physiological stress associated with contrast agent administration, we consider that the breath-holding during the noncontrast scan does not affect the results. The AP image quality results of the two training strategies suggests that the training scheme should have a short duration and a sufficiently long recovery time.

Our goal was to investigate a training strategy that would enhance the AP image quality for patients undergoing their first gadoxetic acid examination. Considering the potential impact of patients’ unfamiliarity with the procedure during the gadoxetic acid-enhanced examination, we specifically selected individuals with no prior experience with gadoxetic acid-enhanced examination. Patients who had previously undergone multiple examinations and achieved success in their first attempt were already familiar with the techniques, which could have introduced more confounding factors if they had also experienced abdominal surgery, postoperative complications, or poor physical condition. Therefore, we excluded data from these patients in this retrospective study to avoid possible confounding.

This study involved certain limitations which should be addressed. First, only two breath-training strategies were included, and additional strategies should be examined to identify the optimal approach. Second, we employed a small, single-center design, and the results should thus be validated in a multicenter study with additional patients. Third, only a single type of scanner (3T) was used, so the results cannot be generalized to other scanners (e.g., 1.5T scanners). Fourth, we were unable to consider the potential influence of patients’ cardiac function and the choice of injection arm (left vs. right antecubital vein) on vascular SI, as this information could not be retrospectively obtained. Future research will address this through a prospective design with a standardized protocol to control for these variables. Fifth, it is customary in our institution to provide patients with breath-holding training before gadoxetic acid-enhanced MRI is conducted. However, the collection of data was limited, as only two nurses maintained consistent habits that could be reviewed, and only four radiographers completely adhered to the internationally recognized injection protocols. This resulted in the exclusion of a substantial amount of patient data. Finally, our suggestion that progressive breathing training may result in differing levels of patient fatigue—which could, in turn, explain the observed difference in TSM incidence but not the similar precontrast scan image quality—is not yet supported by fatigue data, and further validation is necessary. In future studies, we will incorporate respiratory monitoring during MRI scans and fatigue feedback from patient questionnaires to ensure the reliability of our findings. Although the exclusion of a large number of patients might have compromised the integrity of the results by introducing selection bias, we still maintain confidence in the validity of our findings given the extensive cohort that was randomly collected and retrospectively analyzed.

Conclusions

The combination of nongradual breath-hold training and administration of 50% gadoxetic acid was found to be suitable for patients undergoing AP MRI, and our results suggest that this approach can improve AP image quality.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1225/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1225/dss

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1225/coif). X.Z. and M.L. are from Siemens Healthineers China. The other authors have no conflicts of interest to declare.

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. The single-center retrospective study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Board of The First Affiliated Hospital of Army Medical University, Chongqing, China [No. (B)KY2022122] and individual informed consent for this analysis was waived due to the retrospective nature.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Park MJ, Kim YK, Lee MW, Lee WJ, Kim YS, Kim SH, Choi D, Rhim H. Small hepatocellular carcinomas: improved sensitivity by combining gadoxetic acid-enhanced and diffusion-weighted MR imaging patterns. Radiology 2012;264:761-70. [Crossref] [PubMed]
2. Sano K, Ichikawa T, Motosugi U, Sou H, Muhi AM, Matsuda M, Nakano M, Sakamoto M, Nakazawa T, Asakawa M, Fujii H, Kitamura T, Enomoto N, Araki T. Imaging study of early hepatocellular carcinoma: usefulness of gadoxetic acid-enhanced MR imaging. Radiology 2011;261:834-44. [Crossref] [PubMed]
3. Armbruster M, Zech CJ, Sourbron S, Ceelen F, Auernhammer CJ, Rist C, Haug A, Singnurkar A, Reiser MF, Sommer WH. Diagnostic accuracy of dynamic gadoxetic-acid-enhanced MRI and PET/CT compared in patients with liver metastases from neuroendocrine neoplasms. J Magn Reson Imaging 2014;40:457-66. [Crossref] [PubMed]
4. Jeong HT, Kim MJ, Park MS, Choi JY, Choi JS, Kim KS, Choi GH, Shin SJ. Detection of liver metastases using gadoxetic-enhanced dynamic and 10- and 20-minute delayed phase MR imaging. J Magn Reson Imaging 2012;35:635-43. [Crossref] [PubMed]
5. Muhi A, Ichikawa T, Motosugi U, Sou H, Nakajima H, Sano K, Sano M, Kato S, Kitamura T, Fatima Z, Fukushima K, Iino H, Mori Y, Fujii H, Araki T. Diagnosis of colorectal hepatic metastases: comparison of contrast-enhanced CT, contrast-enhanced US, superparamagnetic iron oxide-enhanced MRI, and gadoxetic acid-enhanced MRI. J Magn Reson Imaging 2011;34:326-35. [Crossref] [PubMed]
6. Frydrychowicz A, Jedynak AR, Kelcz F, Nagle SK, Reeder SB. Gadoxetic acid-enhanced T1-weighted MR cholangiography in primary sclerosing cholangitis. J Magn Reson Imaging 2012;36:632-40. [Crossref] [PubMed]
7. Okada M, Murakami T, Kuwatsuru R, Nakamura Y, Isoda H, Goshima S, Hanaoka R, Haradome H, Shinagawa Y, Kitao A, Fujinaga Y, Marugami N, Yuki M, Ichikawa T, Higaki A, Hori M, Fujii S, Matsui O. Biochemical and Clinical Predictive Approach and Time Point Analysis of Hepatobiliary Phase Liver Enhancement on Gd-EOB-DTPA-enhanced MR Images: A Multicenter Study. Radiology 2016;281:474-83. [Crossref] [PubMed]
8. Motosugi U, Ichikawa T, Sou H, Sano K, Tominaga L, Kitamura T, Araki T. Liver parenchymal enhancement of hepatocyte-phase images in Gd-EOB-DTPA-enhanced MR imaging: which biological markers of the liver function affect the enhancement? J Magn Reson Imaging 2009;30:1042-6. [Crossref] [PubMed]
9. Bluemke DA, Sahani D, Amendola M, Balzer T, Breuer J, Brown JJ, Casalino DD, Davis PL, Francis IR, Krinsky G, Lee FT Jr, Lu D, Paulson EK, Schwartz LH, Siegelman ES, Small WC, Weber TM, Welber A, Shamsi K. Efficacy and safety of MR imaging with liver-specific contrast agent: U.S. multicenter phase III study. Radiology 2005;237:89-98. [Crossref] [PubMed]
10. Katsube T, Okada M, Kumano S, Hori M, Imaoka I, Ishii K, Kudo M, Kitagaki H, Murakami T. Estimation of liver function using T1 mapping on Gd-EOB-DTPA-enhanced magnetic resonance imaging. Invest Radiol 2011;46:277-83. [Crossref] [PubMed]
11. Kim YK, Lee MW, Lee WJ, et al. Diagnostic accuracy and sensitivity of diffusion-weighted and of gadoxetic acid-enhanced 3-T MR imaging alone or in combination in the detection of small liver metastasis (≤ 1.5 cm in diameter). Invest Radiol 2012;47:159-66. [Crossref] [PubMed]
12. Davenport MS, Caoili EM, Kaza RK, Hussain HK. Matched within-patient cohort study of transient arterial phase respiratory motion-related artifact in MR imaging of the liver: gadoxetate disodium versus gadobenate dimeglumine. Radiology 2014;272:123-31. [Crossref] [PubMed]
13. Davenport MS, Viglianti BL, Al-Hawary MM, Caoili EM, Kaza RK, Liu PS, Maturen KE, Chenevert TL, Hussain HK. Comparison of acute transient dyspnea after intravenous administration of gadoxetate disodium and gadobenate dimeglumine: effect on arterial phase image quality. Radiology 2013;266:452-61. [Crossref] [PubMed]
14. Goodwin MD, Dobson JE, Sirlin CB, Lim BG, Stella DL. Diagnostic challenges and pitfalls in MR imaging with hepatocyte-specific contrast agents. Radiographics 2011;31:1547-68. [Crossref] [PubMed]
15. Haradome H, Grazioli L, Tsunoo M, Tinti R, Frittoli B, Gambarini S, Morone M, Motosugi U, Colagrande S. Can MR fluoroscopic triggering technique and slow rate injection provide appropriate arterial phase images with reducing artifacts on gadoxetic acid-DTPA (Gd-EOB-DTPA)-enhanced hepatic MR imaging? J Magn Reson Imaging 2010;32:334-40. [Crossref] [PubMed]
16. Poetter-Lang S, Dovjak GO, Messner A, Ambros R, Polanec SH, Baltzer PAT, Kristic A, Herold A, Hodge JC, Weber M, Bastati N, Ba-Ssalamah A. Influence of dilution on arterial-phase artifacts and signal intensity on gadoxetic acid-enhanced liver MRI. Eur Radiol 2023;33:523-34. [Crossref] [PubMed]
17. Tamada T, Ito K, Yoshida K, Kanki A, Higaki A, Tanimoto D, Higashi H. Comparison of three different injection methods for arterial phase of Gd-EOB-DTPA enhanced MR imaging of the liver. Eur J Radiol 2011;80:e284-8. [Crossref] [PubMed]
18. Chung SH, Kim MJ, Choi JY, Hong HS. Comparison of two different injection rates of gadoxetic acid for arterial phase MRI of the liver. J Magn Reson Imaging 2010;31:365-72. [Crossref] [PubMed]
19. Kromrey ML, Tamada D, Johno H, Funayama S, Nagata N, Ichikawa S, Kühn JP, Onishi H, Motosugi U. Reduction of respiratory motion artifacts in gadoxetate-enhanced MR with a deep learning-based filter using convolutional neural network. Eur Radiol 2020;30:5923-32. [Crossref] [PubMed]
20. Gruber L, Rainer V, Plaikner M, Kremser C, Jaschke W, Henninger B. CAIPIRINHA-Dixon-TWIST (CDT)-VIBE MR imaging of the liver at 3.0T with gadoxetate disodium: a solution for transient arterial-phase respiratory motion-related artifacts? Eur Radiol 2018;28:2013-21. [Crossref] [PubMed]
21. Li H, Xiao Y, Wang S, Li Y, Zhong X, Situ W, Xiao E, Zhang Z. TWIST-VIBE five-arterial-phase technology decreases transient severe motion after bolus injection of Gd-EOB-DTPA. Clin Radiol 2017;72:800.e1-6. [Crossref] [PubMed]
22. Kim YK, Lin WC, Sung K, Raman SS, Margolis D, Lim Y, Gu S, Lu D. Reducing Artifacts during Arterial Phase of Gadoxetate Disodium-enhanced MR Imaging: Dilution Method versus Reduced Injection Rate. Radiology 2017;283:429-37. [Crossref] [PubMed]
23. Polanec SH, Bickel H, Baltzer PAT, Thurner P, Gittler F, Hodge JC, Bashir MR, Ba-Ssalamah A. Respiratory motion artifacts during arterial phase imaging with gadoxetic acid: Can the injection protocol minimize this drawback? J Magn Reson Imaging 2017;46:1107-14. [Crossref] [PubMed]
24. Iyama A, Nakaura T, Iyama Y, Kidoh M, Nagayama Y, Oda S, Utsunomiya D, Namimoto T, Morita K, Yuba K, Yamashita Y. Spiral flow-generating tube for saline chaser improves aortic enhancement in Gd-EOB-DTPA-enhanced hepatic MRI. Eur Radiol 2019;29:2009-16. [Crossref] [PubMed]
25. Gutzeit A, Matoori S, Froehlich JM, von Weymarn C, Reischauer C, Kolokythas O, Goyen M, Hergan K, Meissnitzer M, Forstner R, Soyka JD, Doert A, Koh DM. Reduction in respiratory motion artefacts on gadoxetate-enhanced MRI after training technicians to apply a simple and more patient-adapted breathing command. Eur Radiol 2016;26:2714-22. [Crossref] [PubMed]
26. Motosugi U, Ichikawa T, Sou H, Sano K, Ichikawa S, Tominaga L, Araki T. Dilution method of gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI). J Magn Reson Imaging 2009;30:849-54. [Crossref] [PubMed]
27. Poetter-Lang S, Ambros R, Messner A, Kristic A, Hodge JC, Bastati N, Schima W, Chernyak V, Bashir MR, Ba-Ssalamah A. Are dilution, slow injection and care bolus technique the causal solution to mitigating arterial-phase artifacts on gadoxetic acid-enhanced MRI? A large-cohort study. Eur Radiol 2024;34:5215-27. [Crossref] [PubMed]
28. Westbrook JI, Raban MZ, Walter SR, Douglas H. Task errors by emergency physicians are associated with interruptions, multitasking, fatigue and working memory capacity: a prospective, direct observation study. BMJ Qual Saf 2018;27:655-63. [Crossref] [PubMed]
29. Roussos C, Macklem PT. The respiratory muscles. N Engl J Med 1982;307:786-97. [Crossref] [PubMed]
30. Kellmann M. Preventing overtraining in athletes in high-intensity sports and stress/recovery monitoring. Scand J Med Sci Sports 2010;20:95-102. [Crossref] [PubMed]
31. Bishop PA, Jones E, Woods AK. Recovery from training: a brief review: brief review. J Strength Cond Res 2008;22:1015-24. [Crossref] [PubMed]
32. Andersen MB, Williams JM. Athletic injury, psychosocial factors and perceptual changes during stress. J Sports Sci 1999;17:735-41. [Crossref] [PubMed]
33. Lipsman N, Kaping D, Westendorff S, Sankar T, Lozano AM, Womelsdorf T. Beta coherence within human ventromedial prefrontal cortex precedes affective value choices. Neuroimage 2014;85:769-78. [Crossref] [PubMed]
34. Petersen SE, Posner MI. The attention system of the human brain: 20 years after. Annu Rev Neurosci 2012;35:73-89. [Crossref] [PubMed]