T2 mapping of the extraocular muscles in healthy volunteers: preliminary research on scan–rescan and observer

Abstract

Background

T2-mapping technique and derived T2 relaxation time allows quantitative assessment of extraocular muscles; however, the reproducibility of T2 mapping-derived parameters was seldom studied till now.

Purpose

To evaluate the scan–rescan and observer–observer reproducibility of T2 relaxation time measurements of extraocular muscles in young healthy volunteers.

Material and Methods

Fourteen volunteers underwent T2-mapping examinations of the extraocular muscles three times within one month on a 3.0-T MR system. Scan–rescan and observer–observer reproducibility of T2 relaxation time measurements of the extraocular muscles were assessed using intraclass correlation coefficient and coefficient of variation.

Results

Both scan–rescan (short-term and long-term) and observer–observer could achieve good to excellent reproducibility, while better short-term than long-term scan–rescan reproducibility was obtained. The coefficient of variation of the T2 relaxation time of each extraocular muscles during both scan–rescan and observer–observer reproducibility assessment were <6%.

Conclusion

T2 relaxation time measurement of the extraocular muscles is proven to be highly reproducible at 3.0 T. T2 mapping may be a potential imaging technique in the diagnosis and follow-up of orbital diseases involved extraocular muscles in further studies.

Keywords

Extraocular muscles magnetic resonance imaging relaxation observer variation

Introduction

Calculation of T2 relaxation time allows quantitative assessment of investigated organs and structures. An increase in T2 relaxation time usually corresponds to higher water content and can be used as a marker for tissue edema (1). T2 mapping is obtained by acquiring image of the same slice at multiple echoes with a pixel-by-pixel fitting of the T2 relaxation curve (1). Recently, the T2-mapping technique and derived T2 relaxation time have emerged as a potential valuable imaging biomarker for evaluating radiation-induced parotid gland damage, showing cartilage and intervertebral discs degeneration, detecting myocardial edema in acute myocardial infarction and myocarditis, and assessing liver ischemia reperfusion injury (2 –7). In the field of orbital imaging, the T2-mapping technique has also been proven to be promising in detecting active Graves’ ophthalmopathy patients and predicting their response to methylprednisolone treatment (8,9).

However, the reproducibility of quantitative magnetic resonance imaging (MRI)-derived parameters attracts increasing concern. Reproducibility is the agreement between independent results and it reflects biological variation, observer errors, and instrumental errors (10). Knowledge of the reproducibility is crucial for understanding the change of quantitative MRI-derived parameters and its potential influence on its application as an imaging biomarker on disease characterization or treatment response (10). In previous studies, the reproducibility of diffusion-weighted imaging (DWI), intravoxel incoherent motion MRI (IVIM-MRI), and dynamic contrast enhanced MRI (DCE-MRI) have been well studied within various organs (11 –15). Regarding the reproducibility of T2 mapping, few studies had indicated that it was reproducible in musculoskeletal system and myocardium (1,16,17). Therefore, we hypothesized that good reproducibility of the T2-mapping technique could also be obtained when applying in extraocular muscles (EOM).

Therefore, the aim of this study was to preliminarily and prospectively assess the scan–rescan and observer–observer reproducibility of T2 relaxation time measurements of EOMs in young healthy volunteers.

Material and Methods

Study population

This prospective observation study was approved by our institutional ethic committee and written informed consent was obtained from all volunteers. Inclusion criteria were as follows: (i) age >18 years; (ii) no history of any orbital diseases; and (iii) no contraindications for MRI scan. According to these inclusion criteria, we finally enrolled 14 young healthy volunteers (8 men, 6 women; age range = 21–28 years; average age = 23.6 ± 2.0 years) from January 2018 to July 2018.

Imaging acquisition

All volunteers were scanned using a 3.0-T MR scanner (MAGNETOM skyra, Siemens Healthcare, Erlangen, Germany) with a 12-channel head coil. Before MR scan, volunteers were instructed to lie in a supine position and a relaxed status, and to avoid eyeball movement as far as possible. Structural MRI protocol including axial T2-weighted (T2W) imaging (repetition time [TR]/echo time [TE] = 4000/117 ms, slice thickness = 3 mm, slices = 20, field of view [FOV] = 20 cm) and coronal T2W imaging with fat saturation (TR/TE = 4000/75 ms, slice thickness = 3 mm, slices = 18, FOV = 18 cm).

Coronal T2 mapping was obtained using a modified multi-echo spin-echo sequence that implements both model-based accelerated relaxometry by iterative non-linear inversion (MARTINI) (18) and generalized autocalibrating partial parallel acquisition (GRAPPA) (19). This technique was named GRAPPATINI (20). The parameters were as follows: TR/TE = 2020/12–192 ms; delta TE = 12 ms; FOV = 20 cm; slice thickness = 3 mm; slices = 20; echoes = 16; matrix = 279 ×320; GRAPPA acceleration factor = 2; MARTINI acceleration factor = 5. The scan duration was 4 min 40 s. At the end of the first scan (Scan 1), the volunteers were asked to take a rest outside the scanning room. After 30 min, the same T2-mapping sequence was scanned again (Scan 2) for assessing the short-term reproducibility. One week later, the same T2-mapping sequence was scanned again (Scan 3) for assessing the long-term reproducibility.

Imaging analysis

The MRI data sequences were transferred to a Syngo (Leonardo) workstation (Siemens Medical Solution, Erlangen, Germany). T2 relaxation time maps were automatically obtained based on a pixel-wise, mono-exponential, non-negative least squares (NNLS) fit analysis. Two radiologists (observer 1 with six years of experience in head and neck radiology; observer 2 with three years of experience in head and neck radiology) performed the quantitative measurements. Observer 1 measured the data for evaluating the short-term and long-term scan–rescan reproducibility from all three scans. Observer 2 only measured the data from the second scan. The measurement results between observers 1 and 2 were used for evaluating observer–observer reproducibility.

The slice that was 3 mm (one slice) behind the posterior wall of the eyeball was chosen for quantitative measurements of T2 relaxation time. Three circular regions of interest (ROI) (each with an area of 1 mm²) were placed in the central part of the EOMs for the measurements. An average value for each EOM was obtained for statistical analyses. A schematic diagram of the ROI selection methods is shown in Fig. 1.

Fig. 1.

Schematic illustration of region of interest (ROI) selection method. The slice that was 3 mm (one slice) behind the posterior wall of the eyeball was chosen for quantitative measurements (a). With reference to coronal T2-weighted image (b), three circle ROIs (red circle in c) were put on each extraocular muscle (EOM) in the coronal T2 relaxation time map and an average value of each EOM was taken into analysis.

Statistical analysis

All numeric data were reported as mean ± SD; Kolmogorov–Smirnov test was used to analyze the normality of data distribution. One-way analysis of variance (ANOVA) test was used to compare the differences of T2 relaxation time of EOMs among different scans. Independent-samples t-test was used to compare the differences of T2 relaxation time of EOMs between different observers. Short-term and long-term scan–rescan reproducibility and observer–observer reproducibility of the measurements of T2 relaxation time were assessed using the intraclass correlation coefficient (ICC) and coefficient of variation (CV). The agreement was defined as excellent (ICC ≥ 0.81), good (ICC = 0.61–0.80), moderate (ICC = 0.41–0.60), and poor (ICC ≤ 0.40) (14). The CV was defined as the SD of the differences between two measurements divided by the average of both measurements (21). All statistical analyses were performed by using two statistical software packages (SPSS Version 23.0, Chicago, IL, USA; MedCalc Version 11.0, Ostend, Belgium). A two-sided P value <0.05 was considered statistically significant.

Results

Table 1 summarizes the differences of T2 relaxation time of each EOMs among different scans and observers. There were no significant differences on the T2 relaxation time of each EOMs among three scans (P = 0.424–0.964) and between two observers (P = 0.063–0.971).

Table 1.

Differences of T2 relaxation time (ms) of EOMs among different scans and observers.

Parameters	T2 relaxation time (scan–rescan)			P	T2 relaxation time (inter-observer)		P
Parameters	Scan 1	Scan 2	Scan 3	P	Observer 1	Observer 2	P
Right superior rectus	56.562 ± 4.958	56.396 ± 4.279	58.115 ± 4.724	0.565	56.396 ± 4.279	56.718 ± 4.995	0.856
Right inferior rectus	58.857 ± 5.321	60.278 ± 4.831	60.528 ± 5.313	0.656	60.278 ± 4.831	58.016 ± 3.371	0.163
Right lateral rectus	66.462 ± 5.231	66.852 ± 4.710	67.570 ± 3.655	0.811	66.852 ± 4.710	65.077 ± 5.431	0.364
Right medial rectus	57.262 ± 3.726	57.222 ± 4.440	57.558 ± 2.096	0.964	57.222 ± 4.440	56.260 ± 3.039	0.458
Left superior rectus	56.038 ± 3.596	56.912 ± 4.728	56.578 ± 3.037	0.833	56.912 ± 4.728	57.265 ± 4.974	0.849
Left inferior rectus	60.467 ± 4.831	61.350 ± 4.572	61.872 ± 3.656	0.695	61.350 ± 4.572	58.399 ± 3.387	0.063
Left lateral rectus	65.390 ± 5.487	67.514 ± 4.505	65.644 ± 3.736	0.424	67.514 ± 4.505	64.722 ± 3.718	0.085
Left medial rectus	57.133 ± 3.593	57.257 ± 5.449	56.062 ± 5.394	0.778	57.257 ± 5.449	57.323 ± 3.940	0.971

EOM, extraocular muscle.

The scan–rescan reproducibility of the T2 relaxation time of each EOMs was summarized in Table 2. Excellent short-term scan–rescan reproducibility was obtained for the bilateral inferior rectus, bilateral medial rectus, and left superior rectus (ICC = 0.820–0.867), while the bilateral lateral rectus (right: ICC = 0.783, 95% confidence interval [CI] = 0.325–0.930, left: ICC = 0.759, 95% CI = 0.250–0.923) and right superior rectus (ICC = 0.786, 95% CI = 0.334–0.931) showed good scan–rescan reproducibility. By contrast, only the right inferior rectus demonstrated excellent long-term scan–rescan reproducibility (ICC = 0.814, 95% CI = 0.421–0.940), while all the other EOMs showed good long-term scan–rescan reproducibility (ICC = 0.720–0.780). In general, better short-term (ICC = 0.759–0.867) scan–rescan reproducibility was obtained than long-term (ICC = 0.720–0.814). The CVs of the T2 relaxation time of each EOM during both short-term and long-term scan–rescan reproducibility assessments were <6% (short-term range = 1.841–5.372; long-term range = 2.076–5.702). Graphic illustrations of the short-term and long-term scan–rescan variability using Bland–Altman plots are shown in Figs. 2 and 3, respectively.

Table 2.

Scan–rescan and observer–observer reproducibility of T2 relaxation time of each EOM in both eyes.

Parameters	ICC			CV (%)
	Scan–rescan		Observer–observer	Scan–rescan		Observer–observer
	Short term	Long term	Observer–observer	Short term	Long term	Observer–observer
Right superior rectus	0.786 (0.334–0.931)	0.733 (0.167–0.914)	0.775 (0.300–0.928)	3.065	4.565	3.258
Right inferior rectus	0.867 (0.587–0.957)	0.814 (0.421–0.940)	0.816 (0.426–0.941)	3.011	3.238	3.493
Right lateral rectus	0.783 (0.325–0.930)	0.753 (0.229–0.921)	0.760 (0.253–0.923)	3.767	3.980	4.361
Right medial rectus	0.867 (0.584–0.957)	0.763 (0.261–0.924)	0.822 (0.446–0.943)	1.841	2.076	3.417
Left superior rectus	0.824 (0.451–0.943)	0.743 (0.199–0.917)	0.794 (0.357–0.914)	2.943	3.861	2.826
Left inferior rectus	0.849 (0.259–0.952)	0.755 (0.237–0.921)	0.802 (0.385–0.937)	3.052	3.422	4.826
Left lateral rectus	0.759 (0.250–0.923)	0.720 (0.129–0.910)	0.742 (0.197–0.917)	5.372	5.702	3.159
Left medial rectus	0.820 (0.439–0.942)	0.780 (0.316–0.921)	0.823 (0.450–0.943)	3.116	3.339	3.666

Values in parentheses are 95% confidence intervals.

EOM, extraocular muscle; ICC, intra-class correlation coefficient; CV, coefficient of variation.

Fig. 2.

Bland–Altman plots of short-term agreement for the medial rectus (first column), lateral rectus (second column), superior rectus (third column), and inferior rectus (fourth column). The first row indicates the left eye, while the second row indicates the right eye. Differences of T2 values between two measurements of scan 1 and scan 2 (y-axis) were plotted against mean T2 values between two measurements of scan 1 and scan 2 (x-axis), with mean absolute difference (bias) (solid line) and 95% confidence interval of mean difference (limits of agreement) (dashed lines).

Fig. 3.

Bland–Altman plots of long-term agreement for the medial rectus (first column), lateral rectus (second column), superior rectus (third column), and inferior rectus (fourth column). The first row indicates the left eye, while the second row indicates the right eye. Differences of T2 values between two measurements of scan 1 and scan 3 (y-axis) were plotted against mean T2 values between two measurements of scan 1 and scan 3 (x-axis), with mean absolute difference (bias) (solid line) and 95% confidence interval of mean difference (limits of agreement) (dashed lines).

The observer–observer reproducibility of the T2 relaxation time of each EOMs was also summarized in Table 2. Excellent observer–observer reproducibility was achieved for the bilateral inferior and medial rectus (ICC = 0.802–0.823) while good observer–observer reproducibility was achieved for the bilateral superior and lateral rectus (ICC = 0.742–0.794). The CVs of the T2 relaxation time of each EOM during observer–observer reproducibility assessments were <5% (range = 2.826–4.826). Bland–Altman plots of the observer–observer variability are shown in Fig. 4.

Fig. 4.

Bland–Altman plots of observer–observer agreement for the medial rectus (first column), lateral rectus (second column), superior rectus (third column), and inferior rectus (fourth column). The first row indicates the left eye, while the second row indicates the right eye. Differences of T2 values between measurements of two observers (y-axis) were plotted against mean T2 values between measurements of two observers (x-axis), with mean absolute difference (bias) (solid line) and 95% confidence interval of mean difference (limits of agreement) (dashed lines).

Discussion

The normal T2 relaxation time of EOMs reported in our study were in the range of 50.7–72.0 ms, which was similar with the normal value reported in a previous study (9). Hou et al. reported that the mean T2 relaxation time of EOMs was in the range of 68.7–82.8 ms (9). Tiny differences might result from different study cohorts and imaging parameters. Twenty volunteers (mean age = 41.0 ± 14.3 years, age range = 19–62 years) were enrolled in their study, while our study enrolled a younger study cohort (average age = 23.6 ±2.0 years, age range = 21–28 years). In addition, seven TE times (22–88 ms; delta TE = 11 ms) were used in their study, while 16 TE times (12–192 ms; delta TE = 12 ms) were used in our T2 mapping study. In the future, standardization of the imaging parameters is crucial for the application of T2-mapping technique in clinical setting.

Previously, some studies had reported the reproducibility of the T2-mapping technique in other organs (1,16). Roy et al. found that the myocardial T2 value demonstrated high inter-/intra-observer and day-to-day reproducibility at 3.0-T MR (16). Albano et al. insisted that the T2-mapping technique was feasible and reproducible when applied in the sacroiliac joints at 1.5T MR (1). To the best of our knowledge, our study was the first one to evaluate the reproducibility of the T2-mapping technique in EOMs. Satisfactory short-term and long-term scan–rescan and observer–observer agreements found in our study indicated the ability of serially following T2 mapping-derived parameters over time in orbital imaging.

Nemeth et al. compared the immediate (30 min) and short-term (14 days) reproducibility of T2 relaxation time measurements in hip cartilage at 3.0-T MR and found that immediate reproducibility was significantly better than short-term reproducibility (17). Similar results were also found in our study, the short-term (30 min) reproducibility was relatively better than the long-term reproducibility (seven days) (short-term ICC = 0.759–0.867 vs. long-term ICC = 0.720–0.814). Furthermore, as to the measurement of each EOM, we found that the lateral rectus showed relatively lower ICC. It might be associated that our T2-mapping imaging was acquired in a normal coronal plane, not an oblique coronal plane to each orbit. Because of this, the border of the lateral rectus, especially the lateral border, was not as clear as that of the other rectus. Therefore, the measurements on the lateral rectus would fluctuate more obviously than those on the other rectus.

In the clinical setting, due to the association with the early and subtle change of collagen fiber architecture and local water content (1), T2 mapping might allow the detection of the tiny change of the EOMs in various orbital diseases, such as thyroid associated ophthalmopathy (TAO). Derived parameters could help the clinicians to distinguish active from inactive TAO patients, and predict the treatment response to methylprednisolone therapy (9). The good to excellent short-term and long-term scan–rescan reproducibility and observer–observer reproducibility found in our study indicated that T2 mapping was a steady and reliable technique, which could serve as a potential imaging marker in the diagnosis, and especially the follow-up of TAO.

Several limitations should be noticed. First, our study sample size was relatively small. Further study with more participants could provide more power to our conclusion. Second, a previous study by Roy et al. demonstrated that myocardial T2 value decreased significantly with increasing age (16). However, we only included young volunteers in this pilot study. Further correlation investigation between T2 values of EOMs and physical factors (age, gender, body mass index, etc.) would be valuable. Third, magnetic susceptibility artifacts associated with the gas in the maxillary sinus may influence the measurements of the inferior border of the inferior rectus to some extent. Fourth, circular ROIs were used for the measurements which might lead to potential selection bias. Finally, further study applying the T2-mapping technique to assess the EOMs in disease state would be more valuable for clarifying its value on disease staging and therapeutic effect evaluation.

In conclusion, our study indicated that good to excellent scan–rescan and observer–observer reproducibility could be obtained during the measurements of T2 relaxation time of EOMs at 3.0-T MR. T2 mapping was proven to be steady and highly reproducible, and could serve as a useful imaging marker in the diagnosis and follow-up of orbital diseases involving EOMs.

Footnotes

Acknowledgements

The authors thank the members of the Siemens group, Tobias Kober, Yi-Cheng Hsu, and Yi Sun, for their advice and expertise in T2 mapping technology.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Albano

Chianca

Cuocolo

, et al. T2-mapping of the sacroiliac joints at 1.5 Tesla: a feasibility and reproducibility study. Skeletal Radiol 2018; 47:1691–1696.

Zhou

Chu

Dou

, et al. Early evaluation of radiation-induced parotid damage in patients with nasopharyngeal carcinoma by T2 mapping and mDIXON Quant imaging: initial findings. Radiat Oncol 2018; 13:22.

Waldenmeier

Evers

Uder

, et al. Using cartilage MRI T2-mapping to analyze early cartilage degeneration in the knee joint of young professional soccer players. Cartilage 2019; 10:288–298.

Pachowsky

Kleyer

Wagner

, et al. Quantitative T2 mapping shows increased degeneration in adjacent intervertebral discs following kyphoplasty. Cartilage 2018. DOI: 10.1177/1947603518758434.

Xia

Zhang

, et al. Assessment of myocardial edema and area at risk in a rat model of myocardial infarction with a faster T2 mapping method. Acta Radiol 2015; 56:1085–1090.

Baessler

Luecke

Lurz

, et al. Cardiac MRI texture analysis of T1 and T2 maps in patients with infarctlike acute myocarditis. Radiology 2018; 289:357–365.

Hueper

Lang

Hartleben

, et al. Assessment of liver ischemia reperfusion injury in mice using hepatic T2 mapping: Comparison with histopathology. J Magn Reson Imaging 2018; 48:1586–1594.

Majos

Pajak

Grzelak

, et al. Magnetic resonance evaluation of disease activity in Graves’ ophthalmopathy: T2-time and signal intensity of extraocular muscles. Med Sci Monit 2007; 13:44–48.

Hou

, et al. Three dimensional orbital magnetic resonance T2-mapping in the evaluation of patients with Graves’ ophthalmopathy. J Huazhong Univ Sci Technolog Med Sci 2017; 37:938–942.

10.

Sun

, et al. Rectal cancer: Short-term reproducibility of intravoxel incoherent motion parameters in 3.0T magnetic resonance imaging. Medicine (Baltimore) 2017; 96:e6866.

11.

Han

Suo

Sun

, et al. Apparent diffusion coefficient measurement in glioma: Influence of region-of-interest determination methods on apparent diffusion coefficient values, interobserver variability, time efficiency, and diagnostic ability. J Magn Reson Imaging 2017; 45:722–730.

12.

Pan

Zhang

Man

, et al. Measurement and scan reproducibility of parameters of intravoxel incoherent motion in renal tumor and normal renal parenchyma: a preliminary research at 3.0 T MR. Abdom Radiol (NY) 2018; 43:1739–1748.

13.

Jiang

Yin

Cui

, et al. Lung Cancer: Short-term reproducibility of intravoxel incoherent motion parameters and apparent diffusion coefficient at 3T. J Magn Reson Imaging 2018; 47:1003–1012.

14.

Qian W, Xu XQ, Hu H, et al. Dynamic contrast-enhanced MRI in orbital lymphoproliferative disorders: effects of region of interest selection methods on time efficiency, measurement reproducibility, and diagnostic ability. J Magn Reson Imaging 2018;47:1298–1305.

15.

Wang

, et al. Dynamic contrast-enhanced MR imaging in renal cell carcinoma: reproducibility of histogram analysis on pharmacokinetic parameters. Sci Rep 2016; 6:29146.

16.

Roy

Slimani

de Meester

, et al. Age and sex corrected normal reference values of T1, T2 T2* and ECV in healthy subjects at 3T CMR. J Cardiovasc Magn Reson 2017; 19:72.

17.

Nemeth

Di Marco

Boutitie

, et al. Reproducibility of in vivo magnetic resonance imaging T1 rho and T2 relaxation time measurements of hip cartilage at 3.0T in healthy volunteers. J Magn Reson Imaging 2018; 47:1022–1033.

18.

Sumpf

Uecker

Boretius

, et al. Model-based nonlinear inverse reconstruction for T2 mapping using highly undersampled spin-echo MRI. J Magn Reson Imaging 2011; 34:420–428.

19.

Griswold

Jakob

Heidemann

, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002; 47:1202–1210.

20.

Hilbert

Sumpf

Weiland

, et al. Accelerated T2 mapping combing parallel MRI and model-based reconstruction: GRAPPATINI. J Magn Reson Imaging 2018; 48:359–368.

21.

Kamphuis

van der Palen

RLF

de Koning

PJH

, et al. In-scan and scan-rescan assessment of LV in- and outflow volumes by 4D flow MRI versus 2D planimetry. J Magn Reson Imaging 2018; 47:511–522.

T2 mapping of the extraocular muscles in healthy volunteers: preliminary research on scan–rescan and observer–observer reproducibility

Abstract

Background

Purpose

Material and Methods

Results

Conclusion

Keywords

Introduction

Material and Methods

Study population

Imaging acquisition

Imaging analysis

Statistical analysis

Results

Discussion

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

References