Sage Journals: Discover world-class research

Abstract

Objective:

Echo intensity measurements are highly influenced by ultrasound system and parameters used for measurement, making comparisons of results obtained from different ultrasound machines difficult. Therefore, it is necessary to understand how reliability changes when using different ultrasound systems and parameters.

Materials and Methods:

ALOKA SSD4000 and GE LOGIQ P6 systems were used to compare rectus femoris echo intensity in 16 healthy young subjects (eight women) using different depths (D), gains (G), and frequencies (F). The following settings were adopted: ALOKA₁ (D6/G30/F7.5), ALOKA₂ (D6/G45/F7.5), ALOKA₃ (D6/G30/F10), LOGIQ₁ (D6/G50/F15), LOGIQ₂ (D6/G0/F15), LOGIQ₃ (D6/G0/F10), and LOGIQ₄ (D6/G30/F10). Intraclass correlation coefficient, standard error of the measure, minimum difference, and Bland-Altman tests were performed to calculate reliability and agreement between systems’ settings.

Results:

ALOKA₁ × LOGIQ₁, ALOKA₁ × LOGIQ₄, and ALOKA₃ × LOGIQ₁ showed moderate to high ICCs and agreement on the Bland-Altman test.

Conclusion:

Echo intensity varies between systems and parameters, but reliability can be increased by adjusting the ultrasound settings.

Keywords

gray-scale muscle quality reliability frequency gain

Ultrasound (US) measurements of muscle echo intensity (EI) have been widely used in the literature as a measurement of muscle quality.^1,2 EI values have been used to evaluate the effects of aging,^3–6 neuromuscular diseases,^7–9 strength training adaptations,^10,11 and exercise-induced muscle damage.^12–15 This technique allows us to quantify an image’s shades of gray, which in turn allows the identification of the presence of connective tissue, intramuscular fat, or edema, which makes the image whiter (i.e. lower muscle quality), or healthy muscle tissue, which makes the image darker (i.e. higher muscle quality).¹

Significant positive associations have been found in different studies between EI values and isometric peak torque, rate of force development, functional tasks such as 30-second sit-to-stand and preferred gait speed in the elderly,^2,16 sprinting performance in healthy young individuals,¹⁷ stair-climbing performance in professional firefighters,¹⁸ and agility in middle school boys.¹⁹ On the other hand, negative associations have been reported between EI values and exercise capacity in patients with heart failure.²⁰ This is evidence that EI can be used as an important noninvasive tool to study functionality and performance.

Changes in parameters such as depth, gain, and frequency can be used to generate excellent quality measurements. However, changing these parameters can affect the image’s shades of gray, thereby affecting the EI values and muscle quality results. Depth (also known as range), gain (time gain compensation [TGC]), and frequency (rate of ultrasound propagation) play a crucial role in the image’s shades of gray and consequently in muscle EI values.

For a measurement to be reliable, it is essential to know which factors affect it. This allows for controlling intervening variables and ascertaining whether changes in the participant’s results are truly due to an adaptation and not to changes in the US system parameters. Reliability studies have been conducted to quantify the influence of moments,^21,22 raters,^23,24 transducer tilt,²⁵ muscle site,²⁴ image analysis technique,^26,27 and analysts^24,28 on the results, finding that reliability of EI measurements are high to very high. However, few studies have evaluated the influence of using different US systems on muscle EI values. Pillen et al.²⁹ used a phantom (a combination of pig muscles with known shades of gray) and seven healthy subjects to elaborate a conversion equation. This allowed them to obtain EI normal values with the Sonos 2000 Phased Array Imaging System (Hewlett-Packard, Andover, MA) and apply them to measurements made by the Philips IU22 (Philips Healthcare, Eindhoven, The Netherlands). The results helped to improve this technique in clinical practice. Meanwhile, O’Brien et al.³⁰ have compared EI measurements between four different machine/transducer combinations (two machines and two transducers). The US machines used did not have post-processing filters and did not allow changes in the parameters. They found a high reliability in their comparisons; however, the two systems used were of identical models. No study was found in the literature focused on determining the EI reliability, based on measurements, obtained from two different US systems.

Because US is a more accessible imaging tool in clinical practice than magnetic resonance imaging (MRI) and computed tomography, it is in increasing demand, resulting is an increasing variety of US systems available on the market. Thus, clinical studies conducted in different locations and conducted with different US systems. Normally, commercial US systems have built-in postprocessing filters, with the goal of enhancing image resolution and identifying individual structures. However, these filters alter the images’ shades of gray and can affect the EI results.

Because published studies often discuss their results while comparing the values found with those available in the literature,^10,16 it is important to identify the influence of using different US systems when determining the muscle’s EI. Therefore, the aim of the present study was to verify the reliability and the agreement of rectus femoris (RF) muscle EI values calculated from images of a similar muscle location, using two different US systems.

Materials and Methods

This study was conducted according the Declaration of Helsinki, and the local Institutional Research Ethics Committee approved all procedures (project number 708.362). All participants were informed of the benefits and risks prior to signing an institutionally approved consent document for study participation.

Sample size was determined using mean and standard deviation data from a study that evaluated the same variable in a similar population.²² Sixteen healthy young subjects (eight men and eight women; age, 27.56 ± 1.71 years; body mass index [BMI], 23.64 ± 2.79 kg/m²) volunteered to participate in the study. All US images were obtained on the same day while subjects were laying down on a stretcher in a supine position with their knees fully extended. Prior to assessment, volunteers rested for a period of 5 to 10 minutes in order to reestablish body fluids.³¹ Three RF images were obtained from each thigh by the same experienced rater. The transducer from each system was positioned at the same position, at 50% of the thigh length, transversely to the muscle fibers.¹ All images were analyzed on ImageJ software (Version 1.43u; National Institutes of Health, Bethesda, MD) by the same experienced analyst, using a rectangular region of interest that included as much of the muscle as possible.^26,27 Images were obtained using both ALOKA (SSD4000; Aloka, Tokyo, Japan) and LOGIQ (P6; General Electric, Milwaukee, WI) US systems, with all their respective settings. The order in which each system and setting was used was randomized.

ALOKA and LOGIQ US systems were compared using a different combinations of depths (D), gains (G), and frequencies (F). The values were chosen based on a previous qualitative analysis of the images. The values for G were chosen by using the minimum values for both systems (30 for ALOKA and 0 for LOGIQ), plus one intermediate value for ALOKA (45) and two for LOGIQ (30 and 50). The F values chosen for ALOKA were the minimum and maximum available (7.5 and 10), while for LOGIQ, the maximum value and a common value between US systems were chosen (15 and 10, respectively). Depth was set at 6 for both systems. Therefore, for ALOKA, three different settings were adopted: ALOKA₁ (D6/G30/F7.5), ALOKA₂ (D6/G45/F7.5), and ALOKA₃ (D6/G30/F10). For the LOGIQ, four different settings were adopted: LOGIQ₁ (D6/G50/F15), LOGIQ₂ (D6/G0/F15), LOGIQ₃ (D6/G0/F10), and LOGIQ₄ (D6/G30/F10) (Figure 1).

Figure 1.

Images obtained from the same position of the rectus femoris (RF) muscle from a representative subject with both systems and their respective settings: (a) ALOKA₁ (D6/G30/F7.5), (b) ALOKA₂ (D6/G45/F7.5), (c) ALOKA₃ (D6/G30/F10), (d) LOGIQ₁ (D6/G50/F15), (e) LOGIQ₂ (D6/G0/F15), (f) LOGIQ₃ (D6/G0/F10), and (g) LOGIQ₄ (D6/G30/F10).

All ALOKA settings were compared with all LOGIQ settings. An intraclass correlation coefficient (ICC), its 95% confidence interval (CI), standard error of the measure (SEM), and minimum difference (MD) were calculated. This was done to quantify reliability between combinations, reliability between the three images obtained (intrarater reliability), and reliability between images analyzed twice by the same analyst for ten of the study’s participants (intra-analyst reliability). For ICC classification, the following criteria were adopted: no correlation (r = 0.00–0.25), low correlation (r = 0.26–0.49), moderate correlation (r = 0.50–0.69), high correlation (r = 0.70–0.89), and very high correlation (r = 0.90–1.00).³² For moderate and high correlations, a Bland-Altman plot was used to verify RF EI agreement between measurements obtained with the different systems.³³ All analyses were performed with SPSS 20.0 (SPSS, Inc., an IBM Company, Chicago, IL) software package.

Results

Table 1 shows mean and standard deviations for the measures obtained by each US setting. Intrarater reliability was very high (ICC = 0.999; 95% CI, 0.998–0.999; SEM = 0.99 a.u.; MD = 2.76 a.u.). Intra-analyst reliability was also very high (ICC = 0.999; 95% CI, 0.999–0.999; SEM = 0.92 a.u.; MD = 2.55 a.u.). Table 2 shows the reliability results from all the comparisons. ALOKA₁ × LOGIQ₄ and ALOKA₃ × LOGIQ₁ comparisons showed a high correlation (r = 0.700 and 0.736, respectively), while ALOKA₁ × LOGIQ₁ and ALOKA₃ × LOGIQ₄ showed a moderate correlation (r = 0.653 and 0.687, respectively). All other comparisons showed low or no correlation (r = 0.062–0.311).

Table 1.

Mean and Standard Deviation Values for Echo Intensity Analysis Using Different Settings in Both Limbs.

Settings	Right Rectus Femoris (a.u.)		Left Rectus Femoris (a.u.)
Settings	Mean	SD	Mean	SD
ALOKA₁	45.62	17.87	46.38	20.53
ALOKA₂	94.93	27.49	95.39	28.23
ALOKA₃	40.77	18.08	40.42	19.14
LOGIQ₁	42.34	12.98	41.60	11.30
LOGIQ₂	03.69	02.89	04.21	02.85
LOGIQ₃	18.50	07.02	17.41	06.37
LOGIQ₄	51.46	13.15	46.87	12.26

Table 2.

Reliability of the Rectus Femoris Echo Intensity Measurements Obtained From the Two Ultrasonic Systems Using Their Respective Settings.

Comparisons	ICC	95% CI	Correlation	SEM (a.u.)	MD (a.u.)	P Value
ALOKA₁ × LOGIQ₁	0.653	0.302 to 0.829	Moderate	9.34	25.88	.002
ALOKA₁ × LOGIQ₂	0.067	−0.084 to 0.282	No	24.23	67.17	.130
ALOKA₁ × LOGIQ₃	0.236	−0.176 to 0.584	No	17.43	48.32	.012
ALOKA₁ × LOGIQ₄	0.700	0.392 to 0.853	High	8.81	24.42	.001
ALOKA₂ × LOGIQ₁	0.170	−0.128 to 0.497	No	31.02	85.97	.014
ALOKA₂ × LOGIQ₂	0.210	−0.038 to 0.124	No	44.32	122.85	.243
ALOKA₂ × LOGIQ₃	0.062	−0.061 to 0.236	No	42.27	117.18	.062
ALOKA₂ × LOGIQ₄	0.230	−0.160 to 0.583	No	27.56	76.40	.006
ALOKA₃ × LOGIQ₁	0.736	0.457 to 0.871	High	7.90	21.89	.001
ALOKA₃ × LOGIQ₂	0.084	−0.100 to 0.326	No	21.61	59.91	.116
ALOKA₃ × LOGIQ₃	0.311	−0.210 to 0.660	Low	14.77	40.95	.006
ALOKA₃ × LOGIQ₄	0.687	0.280 to 0.855	Moderate	9.08	25.17	.001

Reliability results are expressed by the intraclass correlation coefficient (ICC), 95% confidence interval (CI), standard error of the measure (SEM), minimum difference (MD), and P value.

Bland-Altman plot was used for each of the aforementioned comparisons (ALOKA₁ × LOGIQ₁, ALOKA₁ × LOGIQ₄, ALOKA₃ × LOGIQ₁, and ALOKA₃ × LOGIQ₄) (Figure 2). Bland-Altman plots for ALOKA₁ × LOGIQ₁, ALOKA₁ × LOGIQ₄, and ALOKA₃ × LOGIQ₁ revealed a relatively homogeneous dispersion of EI values within the limits of agreement; however, in the ALOKA₃ × LOGIQ₄, there was a slight positive trend, demonstrating that, for higher EI values (above 40 a.u.), the differences between systems were greater, likely due to higher values obtained from ALOKA.

Figure 2.

Bland-Altman analysis showing the echo intensity agreement of the rectus femoris muscle between the two ultrasonic systems. Dashed lines represent 95% upper and lower limits of agreement (±1.96 SD) and mean difference. Continuous line represents a reference for zero mean difference.

Discussion

The results of the present study demonstrate that EI values are different between US systems, being greatly influenced by settings such as frequency and gain. Therefore, absolute values obtained from different studies should not be compared without establishing normal values for each system, even if the reported parameters for depth, gain, and frequency are similar. It would be advised that results should be reported as a percentage of these values.

By varying the settings of both US systems and comparing results among them, the current study verifies that simply pairing the settings was not enough to make the EI measurements equal between the systems. From the four comparisons that showed a moderate or high correlation, only one had the same settings in both US systems (ALOKA₃ × LOGIQ₄). However, for the same comparison, the Bland-Altman plot did not show agreement between the EI measurements, while all three other comparisons with different parameters did.

A few factors can explain the differences observed in the present study. Each US system has a built-in postprocessing filter that improves image quality, but this limits the uniform shades of gray depicted. Furthermore, the gain adjustment in each system is not the same for each equipment, where, for example, 30% gain in one system does not necessarily correspond to 30% in other systems.

Some strategies can be used to improve the reliability between images obtained in different US systems, making them comparable. Pillen et al.²⁹ used a phantom to obtain a conversion equation that allowed for a reliable use in children. O’Brien et al.³⁰ used a US system without any postprocessing and were able to use the technique reliably. In the present study, researchers experimented with different settings’ combinations, finding combinations in which EI reliability was moderate or high. This allowed the researchers to find a combination of settings that could be used with a satisfactory confidence and without the use of a phantom or of a specific system that possesses no post-processing. The use of a phantom typically adds extra work and a specific system that possesses no post-processing must be specifically designed, as it is not typically found on the market. Using this approach of comparing selected combinations of parameters and finding the most similar ones, hospitals and laboratories that work with distinct US systems can compare their data, if necessary. This could greatly improve their efficiency, resulting in benefits to the patient population.

The EI values obtained did not seem to be influenced either by the rater or by the analyst, as indicated by the very high reliability scores found. However, some limitations of the present study should be acknowledged. The values obtained for different systems and parameters presented large standard deviations, which may have negatively affected the accuracy of the comparisons. The current study has a limited sample of healthy young subjects; therefore, extrapolation of these results should be approached with caution. Finally, the frequencies and gains chosen were limited to just a few for each system, and those numbers were chosen arbitrarily, and settings with different values than ours may help to introduce different combination pairs with higher reliability.

In conclusion, this study analyzed images obtained from two different and popular US systems, using multiple combinations of depth, gain, and frequency. The results indicate that the absolute values obtained from these systems cannot be compared even when the parameters are identical in both systems. However, it was also discovered that there are possible combinations of parameters that render the measurements obtained by the different systems more similar to each other. Comparisons between two US systems, however, should be approached with caution, since even the comparison that obtained the best reliability scores did not present perfect agreement between systems. Further studies should seek to replicate these findings using different US systems and different parameters. This would provide clinicians and researchers with more options to employ different systems when necessary. Another aspect that could greatly benefit clinicians is a possible collaboration between manufacturers. The goal of such a collaboration would be mitigating the differences in images obtained using different US systems.

Footnotes

Declaration of Conflicting Interests

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

Funding

The authors received no financial support for the research,authorship,and/or publication of this article.

ORCID iD

Rodrigo Rabello

References

Fukumoto

Ikezoe

Yamada

, et al: Skeletal muscle quality assessed from echo intensity is associated with muscle strength of middle-aged and elderly persons. Eur J Appl Physiol 2012;112(4):1519–1525.

Lopez

Wilhelm

Rech

Minozzo

Radaelli

Pinto

: Echo intensity independently predicts functionality in sedentary older men. Muscle Nerve 2017;55(1):9–15.

Jacobs

Jansen

Janssen

Raijmann

Van Alfen

Pillen

: Quantitative muscle ultrasound and muscle force in healthy children: a 4-year follow-up study. Muscle Nerve 2013;47(6):856–863.

Nishihara

Kawai

Hayashi

, et al: Frequency analysis of ultrasonic echo intensities of the skeletal muscle in elderly and young individuals. Clin Interv Aging 2014;9:1471–1478.

Arts

IMP

Pillen

Schelhaas

Overeem

Zwarts

: Normal values for quantitative muscle ultrasonography in adults. Muscle Nerve 2010;41(1):32–41.

Scholten

Pillen

Verrips

Zwarts

: Quantitative ultrasonography of skeletal muscles in children: normal values. Muscle Nerve 2003;27(6):693–698.

Pillen

Arts

IMP

Zwarts

: Muscle ultrasound in neuromuscular disorders. Muscle Nerve 2008;37(6):679–693.

Fischer

Carpenter

Hartlage

Carroll

Stephens

: Muscle imaging in neuromuscular disease using computerized real-time sonography. Muscle Nerve 1988;11(3):270–275.

Reimers

Fleckenstein

Witt

Müller-Felber

Pongratz

: Muscular ultrasound in idiopathic inflammatory myopathies of adults. J Neurol Sci 1993;116(1):82–92.

10.

Cadore

González-Izal

Pallarés

, et al: Muscle conduction velocity, strength, neural activity, and morphological changes after eccentric and concentric training. Scand J Med Sci Sport 2014;24(5):e343–e352.

11.

Fukumoto

Tateuchi

Ikezoe

, et al: Effects of high-velocity resistance training on muscle function, muscle properties, and physical performance in individuals with hip osteoarthritis: a randomized controlled trial. Clin Rehabil 2014;28(1):48–58.

12.

Chen

Nosaka

Chen

: Muscle damage protection by low-intensity eccentric contractions remains for 2 weeks but not 3 weeks. Eur J Appl Physiol 2012;112(2):555–565.

13.

Chen

Lin

Chen

Lin

Nosaka

: Comparison in eccentric exercise-induced muscle damage among four limb muscles. Eur J Appl Physiol 2011;111(2):211–223.

14.

Chen

Tseng

Huang

Chen

Tseng

Nosaka

: Low-intensity eccentric contractions attenuate muscle damage induced by subsequent maximal eccentric exercise of the knee extensors in the elderly. Eur J Appl Physiol 2013;113(4):1005–1015.

15.

Nosaka

Newton

Sacco

Chapman

Lavender

: Partial protection against muscle damage by eccentric actions at short muscle lengths. Med Sci Sports Exerc 2005;37(5):746–753.

16.

Rech

Radaelli

Goltz

da Rosa

LHT

Schneider

Pinto

: Echo intensity is negatively associated with functional capacity in older women. Age (Omaha) 2014;36(5):9708.

17.

Mangine

Fukuda

Townsend

, et al: Sprinting performance on the Woodway Curve 3.0TM is related to muscle architecture. Eur J Sport Sci 2015;15(7):606–614.

18.

Kleinberg

Ryan

Tweedell

Barnette

Wagoner

: Influence of lower extremity muscle size and quality on stair-climb performance in career firefighters. J Strength Cond Res 2016;30(6):1613–1618.

19.

Stock

Mota

Hernandez

Thompson

: Echo intensity and muscle thickness as predictors of athleticism and isometric strength in middle-school boys. Muscle Nerve 2017;55(5):685–692.

20.

Nakano

Hori

Fukushima

, et al: Enhanced echo intensity of skeletal muscle is associated with exercise intolerance in patients with heart failure [published online September 15, 2019]. J Card Fail.

21.

Vieira

Siqueira

Ferreira-Junior

Pereira

Wagner

Bottaro

: Ultrasound imaging in women’s arm flexor muscles: intra-rater reliability of muscle thickness and echo intensity. Brazilian J Phys Ther 2016;20(6):535–542.

22.

Santos

Armada-da-Silva

PAS

: Reproducibility of ultrasound-derived muscle thickness and echo-intensity for the entire quadriceps femoris muscle. Radiography 2017;23(3):e51–e61.

23.

Zaidman

Wilder

Darras

Rutkove

: Minimal training is required to reliably perform quantitative ultrasound of muscle. Muscle Nerve 2014;50(1):124–128.

24.

Rabello

Fröhlich

Bueno

, et al: Echo intensity reliability between two rectus femoris probe sites. Ultrasound 2019;27(4):233–240.

25.

Ishida

Suehiro

Suzuki

Watanabe

: Muscle thickness and echo intensity measurements of the rectus femoris muscle of healthy subjects: intra and interrater reliability of transducer tilt during ultrasound. J Bodyw Mov Ther 2018;22(3):657–660.

26.

Lanferdini

Manganelli

Lopez

Klein

Cadore

Vaz

: Echo intensity reliability for the analysis of different muscle areas in athletes. J Strength Cond Res 2019;33(12):3353–3360.

27.

Caresio

Molinari

Emanuel

Minetto

: Muscle echo intensity: reliability and conditioning factors. Clin Physiol Funct Imaging 2015;35(5):393–403.

28.

Sarwal

Parry

Berry

, et al: Interobserver reliability of quantitative muscle sonographic analysis in the critically ill population. J Ultrasound Med 2015;34(7):1191–1200.

29.

Pillen

Van Dijk

Weijers

Raijmann

De Korte

Zwarts

: Quantitative gray-scale analysis in skeletal muscle ultrasound: a comparison study of two ultrasound devices. Muscle Nerve 2009;39(6):781–786.

30.

O’brien

Cazares Gonzalez

Ghosh

Mandrekar

Boon

: Reliability of a novel ultrasound system for gray-scale analysis of muscle. Muscle Nerve 2017;56(3):408–412.

31.

Lopez

Pinto

: Does rest time before ultrasonography imaging affect quadriceps femoris muscle thickness, cross-sectional area and echo intensity measurements? Ultrasound Med Biol 2019;45(2):612–616.

32.

Munro

Statistical Methods for Health Care Research. Philadelphia, PA, Lippincott Williams & Wilkins, 1986.

33.

Bland

Altman

. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1(8476):307–310.

Echo Intensity Reliability From Two Ultrasound Systems

Abstract

Objective:

Materials and Methods:

Results:

Conclusion:

Keywords

Materials and Methods

Results

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iD

References