Variability and Reproducibility of Segmental Longitudinal Strain Measurement

JACC: CARDIOVASCULAR IMAGING

VOL.

ª 2017 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER

-, NO. -, 2017

ISSN 1936-878X/$36.00 http://dx.doi.org/10.1016/j.jcmg.2017.01.027

Variability and Reproducibility of Segmental Longitudinal Strain Measurement A Report From the EACVI-ASE Strain Standardization Task Force Oana Mirea, MD, PHD,a Efstathios D. Pagourelias, MD, PHD,a Jurgen Duchenne, MSC,a Jan Bogaert, MD, PHD,b James D. Thomas, MD,c Luigi P. Badano, MD, PHD,d Jens-Uwe Voigt, MD, PHD,a on behalf of the EACVI-ASE-Industry Standardization Task Force

ABSTRACT OBJECTIVES In this study, we compared left ventricular (LV) segmental strain measurements obtained with different ultrasound machines and post-processing software packages. BACKGROUND Global longitudinal strain (GLS) has proven to be a reproducible and valuable tool in clinical practice. Data about the reproducibility and intervendor differences of segmental strain measurements, however, are missing. METHODS We included 63 volunteers with cardiac magnetic resonance–proven infarct scar with segmental LV function ranging from normal to severely impaired. Each subject was examined within 2 h by a single expert sonographer with machines from multiple vendors. All 3 apical views were acquired twice to determine the test-retest and the intervendor variability. Segmental longitudinal peak systolic, end-systolic, and post-systolic strain were measured using 7 vendor-specific systems (Hitachi, Tokyo, Japan; Esaote, Florence, Italy; GE Vingmed Ultrasound, Horten, Norway; Philips, Andover, Massachusetts; Samsung, Seoul, South Korea; Siemens, Mountain View, California; and Toshiba, Otawara, Japan) and 2 independent software packages (Epsilon, Ann Arbor, Michigan, and TOMTEC, Unterschleissheim, Germany) and compared among vendors. RESULTS Image quality and tracking feasibility differed among vendors (analysis of variance, p < 0.05). The absolute test-retest difference ranged from 2.5% to 4.9% for peak systolic, 2.6% to 5.0% for end-systolic, and 2.5% to 5.0% for post-systolic strain. The average segmental strain values varied significantly between vendors (up to 4.5%). Segmental strain parameters from each vendor correlated well with the mean of all vendors (r2 range 0.58 to 0.81) but showed very different ranges of values. Bias and limits of agreement were up to
4.6  7.5%. CONCLUSIONS In contrast to GLS, LV segmental longitudinal strain measurements have a higher variability on top of the known intervendor bias. The fidelity of different software to follow segmental function varies considerably. We conclude that single segmental strain values should be used with caution in the clinic. Segmental strain pattern analysis might be a more robust alternative. (J Am Coll Cardiol Img 2017;-:-–-) © 2017 by the American College of Cardiology Foundation.

From the aDepartment of Cardiovascular Diseases, University Hospital Leuven, Leuven, Belgium; bDepartment of Radiology, University Hospital Leuven, Leuven, Belgium; cBluhm Cardiovascular Institute, Northwestern University, Chicago, Illinois; and the

d

Cardiac, Thoracic and Vascular Sciences, University Padua, Padua, Italy. Dr. Mirea is permanently affiliated to the

Department of Cardiology, University Hospital of Craiova, Romania. This study was supported by a dedicated grant from the American Society of Echocardiography. Dr. Mirea has received a research grant from the European Association of Cardiovascular Imaging. Dr. Pagourelias holds a research grant from the European Association of Cardiovascular Imaging. Dr. Thomas has received honoraria and consulting fees from Edwards, Abbott, and GE. Dr. Voigt holds a personal research mandate from the Flemish Research Foundation; and has received a research grant from the University Hospital Gasthuisberg. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received October 21, 2016; revised manuscript received January 24, 2017, accepted January 26, 2017.

2

Mirea et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017 - 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

ABBREVIATIONS AND ACRONYMS ANOVA = analysis of variance AVC = aortic valve closure CMR = cardiac magnetic resonance

ECG = electrocardiogram ES = end-systolic GLS = global longitudinal strain

ICC = intraclass correlation LGE = late gadolinium enhancement

LS = longitudinal strain LV = left ventricular PS = peak systolic PSS = post-systolic strain

T

wo-dimensional echocardiography

speckle has

tracking

been

pro-

posed for improving the echocardio-

good condition, in sinus rhythm, and had no evidence of cardiac disease in their history, resting electrocardiography (ECG), and baseline echocardiogram.

graphic quantification of left ventricular (LV)

The study was approved by the ethical commission

segmental and global function. Longitudinal

of the University Hospitals Leuven and all subjects

strain (LS) appears to be the most robust

gave written informed consent before inclusion.

among the various myocardial strain compo-

INDUSTRY

nents, and global longitudinal strain (GLS)

partners within the task force were invited to partic-

has demonstrated added diagnostic and

ipate in the study by an open letter. Seven ultrasound

prognostic value in a wide range of condi-

machine

tions such as heart failure (1), valvular heart

machine, speckle tracking software, and an applica-

PARTNER

RECRUITMENT. All

manufacturers

provided

an

industry

ultrasound

disease (2), and others. In a previous study

tion specialist to optimize data acquisition for the

by this task force, we showed good reproduc-

study. Additionally, 2 manufacturers of generic

ibility of GLS measurements (3), suggesting

software solutions for speckle tracking analysis

that this technique can be safely used in the

participated in the comparison. One company (Phi-

clinic, in particular for repeated measure-

lips) withdrew later from the study for technical

ments in the same patient (4,5).

reasons. A list of participants is provided in Table 1.

Previous studies have demonstrated that the assessment of segmental LS provides added information in wide range of pathologies (6–9). However, data about the reproducibility of segmental strain are conflicting (10,11), and intervendor differences remain to be assessed. In the context of the ongoing work of the task force on strain standardization, which was initiated by the European Association of Cardiovascular Imaging (EACVI) and the American Society of Echocardiography in collaboration with industry (12,13), we have set up a study to investigate the robustness, reproducibility, and intervendor variability of segmental speckle tracking–based strain measurements.

METHODS

STUDY PROTOCOL. Echocardiographic imaging. The echo-

cardiographic image acquisitions were completed in 5 days during 9 sessions of 2 to 3 h each. Seven subjects were simultaneously scanned with the 7 different ultrasound machines. To minimize differences in image acquisition, 1 experienced examiner (with at least 2 years of experience in routine echocardiography, including the use of speckle tracking) was assigned to each subject and both rotated through all machines. Examiners were responsible for the acquisition of high-quality standard echocardiographic images. In addition, application specialists ensured optimal machine settings and image acquisition according to the respective manufacturers’ recommendations. Blood pressure was measured at the beginning and at the end of the echocardiography session. Patients

STUDY POPULATION. The study population comprised

were examined in left lateral decubitus position. LV

patients with prior myocardial infarction and healthy

4-, 3-, and 2-chamber views were acquired during

volunteers. A short list of 63 potential patients was

breath hold. Pulsed wave Doppler recordings of the

created from hospital records on the basis of the

mitral inflow and aortic outflow were obtained for

following criteria: 1) age >18 years and ability to con-

timing measurements. Examiners then left the ex-

sent, walk, and lie in supine position for 2 hours; 2)

amination bed for at least 1 min and walked around.

good acoustic window and regular heart rhythm; 3) a

After that, a second set of apical views was acquired

documented myocardial infarction within maximum 2

for the assessment of test-retest variability.

years before the study; and 4) the existence of a late

A minimum of 3 consecutive cycles was recorded per

gadolinium enhancement (LGE) cardiac magnetic

view. All image data were stored as raw data in a pro-

resonance (CMR) study performed after the myocar-

prietary company format if available. In addition, all

dial infarction (without other ischemic events or car-

data were also stored in standard Digital Imaging and

diac interventions before the image acquisitions for

Communications in Medicine format to allow post-

this study). Patients were then contacted by telephone

processing with the independent software packages.

and invited to participate in the study. Care was taken

CMR imaging. All CMR studies were performed on a

to cover a wide range of segmental and functional ab-

1.5-T

normalities. In case not all invited patients would

Netherlands). Cine images were taken in horizontal,

present for the study, healthy volunteers were

vertical, and short-axis views. Ten minutes after

recruited as “gap fillers in stand-by” from the co-

intravenous

workers of our imaging laboratory. They were all in

tetraazacyclododecane-tetraacetic

Philips

Intera-CV

bolus

of

0.2

(Philips,

mmol/kg acid

Best,

the

gadolinium(Dotarem,

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017

Mirea et al.

- 2017:-–-

Guerbet,

Villepinte,

Variability of Segmental Longitudinal Strain Measurements

France),

LGE

images

were

acquired in the same views. LGE images were used

T A B L E 1 Vendors Participating in the Study With Type and Version of

Equipment Provided

to verify the existence of scar and to determine the segmental scar burden per patient using an 18-segment model (14). Healthy volunteers were

Vendor

Hitachi

Ultrasound Machine

Prosound f75

Type

High end

Software and Version

2DTT Analysis v6.0

Esaote

MYLABALPHA eHD

Portable

XStrain2D- v5.50

GE

Vivid E9

High end

EchoPac v20.1

DATA ANALYSIS. Conventional echocardiographic parameters.

Philips

Epiq

High end

*

Image quality was scored per vendor and per segment

Samsung

RS80A with Prestige

High end

Kardia 1.00.0615

(Online Appendix). Biplane volumes and ejection

Siemens

Acuson S2000 CV system

High end

syngo VVI v4.0

fraction were calculated by using modified Simpson

Toshiba

Artida

High end

rule (4). The measurements were performed on the

Epsilon†

EchoInsight

TOMTEC†

2D CPA 1.3 (module of TOMTEC ARENA)

assumed to have no scar without CMR examination.

images of 1 vendor (GE). Strain measurements. All strain measurements were performed by a single observer (O.M.). This observer

ACP v3.2

*Prototype software withdrawn for technical reasons. †Software-only vendor.

had a solid background in tissue Doppler and speckle tracking analysis (>2,000 analyses) before starting the data analysis of this study. Images were analyzed using the vendor-specific speckle tracking software. Before the analysis, the observer was trained by application specialists from each company in the use of their software packages. There was no specific order of the vendor analysis. For the 2 independent software pro-

All strain values are reported as measured (i.e., more negative values represent more shortening). Because it is common usage, we refer in the discussion to the measured absolute amount of strain (i.e., “higher strain values” indicates that measured LS values were more negative).

viders, Digital Imaging and Communications in Medi-

STATISTICAL ANALYSIS. Test-retest variability was

cine images acquired with the GE system were used.

assessed by intraclass correlation (ICC) (2-way mixed

With each software, patient data were analyzed in

model, absolute agreement between single measure-

the order of the study identification number. In each

ments) and as absolute error between repeated mea-

view, the cardiac cycle with the best image quality

surements. The bias between software packages in

was selected. End-diastole was manually set to the R

segmental strain was compared by repeated measures

peak of the ECG. If the software did not allow that,

analysis of variance (ANOVA). Further, segmental

the automatic settings of the software were used instead. The aortic valve closure (AVC) was measured from the pulsed wave Doppler recording of the LV

F I G U R E 1 General Pattern of a Longitudinal Strain Curve

outflow tract and AVC was manually set to this time in all software packages. Next, the region of interest was created either by manually tracing the endocardium or by automated recognition according to the requirements of the software. All post-processing settings were maintained as recommended by the vendor. The quality of the tracking was assessed for each segment by visually comparing the tracking result with the underlying myocardial motion. Segments were excluded from further analysis when the tracking did not follow accurately the myocardial motion after at least 2 attempts of re-adjusting the region of interest. We used an 18-segment model (3 segments/wall), according to the recommendations for segmental

In the background, a spectral Doppler flow profile of the aortic valve with a marked closure artifact is shown. Yellow line

function analysis (4,13). In each segmental strain

represents the strain curve. White dashed line indicates the

curve, peak systolic (PS), end-systolic (ES), and post-

timing of the R peak in the electrocardiogram as surrogate of

systolic strain (PSS) peak were measured. The peaks were defined as follows: PS, maximum (positive or negative) strain value before AVC; ES, strain value at AVC; and PSS, the maximum negative deformation after AVC, if more negative than ES (Figure 1).

ED. AVC ¼ aortic valve closure; ED ¼ end-diastole; ES ¼ endsystolic strain, measured at the time of aortic valve closure; PS ¼ peak systolic strain, defined as positive or negative peak during systole; PSS ¼ post-systolic strain, any peak more negative than end-systolic strain after aortic valve closure.

3

Mirea et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017 - 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

F I G U R E 2 Percentage of Segments Excluded Because of Bad Tracking

100 90

Excluded Segments (%)

4

9.5

11.0

8.9

7.1

16.9

15.1

Samsung

Siemens

Toshiba

Epsilon

* * *

* * *

* *

*

22.9

12.5

80 70 60 50 40 30 20 10 0

Hitachi Esaote GE Samsung Siemens Toshiba Epsilon TOMTEC

Esaote

* * * * *

* * * * *

GE

* * * *

Included

TOMTEC

Excluded

Pink indicates bad tracking. Analysis of variance post hoc results for differences between vendors are detailed in the table underneath. *p < 0.05.

measurements of each vendor were compared to the

a total of 882 echocardiographic examinations (2 ex-

average of all vendors for the same segment. Addi-

aminations on 7 machines/subject) could be per-

tionally, ICC coefficients were calculated among

formed. Systolic arterial blood pressure increased

vendors (Online Appendix).

slightly during the scanning session (128  20 mm Hg to 135  17 mm Hg; p < 0.05), whereas diastolic blood

RESULTS

pressure remained unchanged (73  13 to 74  9;

PATIENT CHARACTERISTICS. Of 63 patients initially

invited to the study, 5 dropped out (3 no-shows, 1 atrial fibrillation, 1 physical inability to complete all echocardiographic examinations) and had to be replaced by healthy volunteers. Patient characteristics are summarized in Online Table 1. As planned,

p ¼ 0.6). The ejection fraction in our study population ranged from 28% to 73% (average 52.4  9.9%). Segmental scar burden could be defined in all 1,134 segments. Of these, 748 (66%) had no evidence of scar, 129 (11.4%) had a nontransmural scar, 241 (21.3%) a transmural scar (>75% of wall thickness), and only 16 (1.4%) were partially scarred (20% to 80% of segment length).

T A B L E 2 Test-Retest Agreement: ICCs of PS, ES, and PSS With 95% CIs

Vendor

TRACKING FEASIBILITY. The number of segments

that could not be tracked with acceptable quality

PS (95% CI)

ES (95% CI)

PSS (95% CI)

Hitachi

0.78 (0.75–0.86)

0.79 (0.76–0.81)

0.78 (0.76–0.81)

differed significantly between vendors, ranging from

Esaote

0.78 (0.75–0.81)

0.76 (0.73–0.78)

0.71 (0.68–0.74)

7.1% to 22.9% (ANOVA p < 0.05) (Figure 2). Wall-by-

GE

0.90 (0.89–0.91)

0.90 (0.89–0.91)

0.88 (0.86–0.89)

wall analysis revealed that the interventricular

Samsung

0.77 (0.73–0.79)

0.75 (0.71–0.78)

0.70 (0.67–0.74)

septum and the inferior wall had the highest tracking

Siemens

0.72 (0.69–0.76)

0.71 (0.67–0.74)

0.67 (0.63–0.70)

Toshiba

0.86 (0.84–0.87)

0.87 (0.85–0.88)

0.83 (0.81–0.85)

Epsilon

0.80 (0.77–0.82)

0.79 (0.77–0.82)

0.74 (0.71–0.77)

TOMTEC

0.80 (0.78–0.82)

0.80 (0.77–0.82)

0.78 (0.75–0.80)

CI ¼ confidence interval; ES ¼ end-systolic strain; ICC ¼ intraclass correlation coefficient; PS ¼ peak systolic strain; PSS ¼ post-systolic strain.

feasibility while the anterior wall segments were most difficult to track. Online Figure 1 shows the exclusion per segment. TEST-RETEST VARIABILITY OF SEGMENTAL STRAIN MEASUREMENTS. The test-retest agreement ranged

from moderate to excellent (ICC coefficients: 0.67 to

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017

Mirea et al.

- 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

Average Absolute Test-re-test Difference (%)

F I G U R E 3 Average Absolute Test-Retest Difference of Segmental Longitudinal PS, ES, and PSS Measurements for Each Vendor

12

10

8

6

4

2

0

Hitachi

Esaote

GE

Samsung

Siemens

Toshiba

Epsilon

TOMTEC

PS

3.8

3.6

3.0

4.4

4.9

2.6

4.0

4.4

ES

3.7

3.6

2.9

6.4

5.0

2.6

4.0

4.4

PSS

3.5

3.5

2.8

4.2

5.0

2.5

4.0

4.3

The outlier in Samsung ES strain can be explained by the used prototype software, which had no function for a reliable AVC measurement. Abbreviations as in Figure 1.

0.90) (Table 2) and showed significant differences between vendors (ANOVA p < 0.05).

Table 4 shows the Pearson correlation coefficients for the pairwise vendor vs. vendor comparisons of PS

The average absolute difference between LS values

values. Data for ES and PSS are provided in Online

from the same segment in the first and second image

Tables 5 and 6. The Bland-Altman analysis of the

acquisition ranged from 2.6% to 4.9% for PS, 2.6% to

same comparisons is provided in Table 5 and Online

5.0% for ES, and 2.5% to 5.0% for PSS (Figure 3).

Tables 7 and 8.

Interestingly, the test/retest variability of PS, ES, and

The correlation of segmental PS strain measure-

PSS was not significantly different within a given

ments from a given vendor with the segmental mean of

vendor, except for Samsung, where it was higher for

all vendors ranged from r2 ¼ 0.58 to r2 ¼ 0.81 (Figure 5).

ES (ANOVA p > 0.05 and p < 0.01, respectively).

The slopes of the regression lines reveal that the range

Online Table 2 shows the absolute difference per

of measured strain values differs among vendors with

level and per segment for PS. In general, midwall

GE having the highest range (slope: 1.29) and Esaote

segments had the lowest test-retest variability in all

having the lowest (slope: 0.82) (Figure 5). The analysis

vendors.

per ventricular level (apical, mid, basal) revealed that

INTERVENDOR DIFFERENCES. The average values of

the basal segments showed the lowest correlation with

PS, ES, and PSS LS of all segments of our study cohort

the mean of all vendors (r 2 from 0.50 to 0.69), whereas

are displayed in Figure 4. The maximum absolute

the apical segments showed the highest (r 2 from 0.76 to

difference between the vendors with the highest and

0.85) (Online Figure 2). The same analysis reveals that

lowest values was 4.5% for all 3 parameters. In more

the differences in the slope of the regression line are

than one-half of the post hoc comparisons, the bias

most pronounced in the apex.

between vendors reached statistical significance (ANOVA p < 0.05) (Figure 4).

DISCUSSION

The intervendor agreement of LV segmental PS values between vendors ranged from poor to good (ICC

MAIN FINDINGS OF THE STUDY. In this study, we

between 0.52 and 0.79, Table 3). The intervendor

directly

agreement of LV segmental ES was similar (ICC be-

segmental strain measurements from 6 ultrasound

compared

the

speckle

tracking–based

tween 0.52 and 0.79, Online Table 3) and slightly lower

machine vendors and 2 software-only vendors in a

for PSS (ICC between 0.45 and 0.77, Online Table 4).

group of volunteers with myocardial segmental

5

Mirea et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017 - 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

F I G U R E 4 Average Segmental PS, ES, and PSS Values per Vendor

-30

-25

-20

Strain (%)

6

-15

-10

-5

0

Hitachi

Esaote

GE

Samsung

Siemens

Toshiba

Epsilon

TOMTEC

PS

-13.3

-16.9

-17.8

-15.0

-17.0

-16.6

-16.9

-15.7

ES

-13.3

-16.3

-17.8

-14.9

-16.6

-16.5

-16.8

-15.7

PSS

-15.0

-17.0

-19.5

-16.7

-18.3

-17.5

-18.3

-18.0

Esaote GE Samsung Siemens Toshiba Epsilon TOMTEC

* * * * * * *

* * * * * * *

* * * * * * * * * * * * * * * * * * * * * * * * * * * *

*Analysis of variance post hoc test with p < 0.05. Abbreviations as in Figure 1.

function ranging from normal to severely impaired.

The specific algorithms used by speckle tracking

We found that: 1) the feasibility of assessing segmental

software solutions from different vendors may have a

strain differs significantly among vendors; 2) the test-

significant impact on strain results. We have there-

retest variability is relatively high but has also a

fore aimed particularly at finding direct or indirect

considerable range among vendors; 3) the intervendor

evidence for differences in the processing of the data.

bias is relevant; and 4) the range of measured segmental strain values differs between vendors.

Most of the speckle tracking algorithms apply noise reduction by temporal and spatial smoothing. It can be expected that extensive smoothing improves the

MEASUREMENT VARIABILITY: SELECTED ASPECTS.

robustness of GLS assessments but may lead to a

A more extensive discussion of potential sources of

lower sensitivity towards small segmental or tempo-

measurement variability is provided in the Online

ral abnormalities. In our comparisons, we found a

Appendix.

considerable difference in the range of measured

T A B L E 3 ICCs and 95% CIs of PS Measurement Among All Vendors

Hitachi

Esaote

Esaote

GE

Samsung

Siemens

Toshiba

Epsilon

0.52 (0.34–0.64)

GE

0.55 (0.30–0.69)

Samsung

0.62 (0.56–0.67)

0.54 (0.48–0.60)

0.71 (0.59–0.78)

Siemens

0.57 (0.37–0.70)

0.59 (0.54–0.63)

0.70 (0.66–0.73)

0.65 (0.59–0.71)

Toshiba

0.64 (0.43–0.76)

0.68 (0.64–0.71)

0.72 (0.67–0.76)

0.65 (0.61–0.70)

0.66 (0.62–0.70)

Epsilon

0.54 (0.39–0.64)

0.55 (0.50–0.59)

0.79 (0.76–0.82)

0.69 (0.64–0.74)

0.64 (0.60–0.67)

0.70 (0.66–0.73)

TOMTEC

0.61 (0.54–0.66)

0.59 (0.54–0.63)

0.71 (0.62–0.77)

0.63 (0.59–0.67)

0.67 (0.61–0.71)

0.68 (0.64–0.71)

Abbreviations as in Table 2.

0.59 (0.54–0.63)

0.67 (0.63–0.71)

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017

Mirea et al.

- 2017:-–-

7

Variability of Segmental Longitudinal Strain Measurements

segmental stain values among different vendors. A high range of values could theoretically be due to

T A B L E 4 Pearson Correlation Coefficients for PS Measurements Among All Vendors

Hitachi

high noise levels; however, because the correlation

Esaote

GE

Samsung

Siemens

Toshiba

between vendors was acceptable, it must be assumed

Esaote

that the range of measured strain values from a

GE

0.645

0.616

Samsung

0.644

0.564

Siemens

0.645

0.603

0.699

0.670

Toshiba

0.699

0.678

0.775

0.679

function are (i.e., how much spatial smoothing is

Epsilon

0.586

0.552

0.804

0.705

0.638

0.711

applied to the data). A higher range of values would

TOMTEC

0.637

0.614

0.732

0.631

0.680

0.716

respective software is rather reflecting how rigid its assumptions about the homogeneity of myocardial

Epsilon

0.575 0.743 0.683 0.684

than rather reflect the better fidelity of a software in Abbreviation as in Table 2.

following myocardial motion locally. Strain measurements rely strongly on the definition of cardiac time events (15). An accurate identification of end-diastole and ES is of particular importance in segmental disease when the timing of strain peaks becomes as important as the amplitude. In the noncommercial Samsung software, where options for manual setting of AVC were limited because of a preliminary user interface, we consequently found a clearly higher variability in ES strain measurements which depend most on the accurate definition of AVC (Figures 1 and 3). cardium and lowest in the epicardium (16). It is important

to

consider

the segmental strain analysis was performed on a larger number of segments, and the results showed poor intraobserver and interobserver reproducibility (11). A more recent study, which reported reproducibility data of PS strain, showed very good test-retest agreement (ICC ranged from 0.88 to 0.97) (10). It is not clear, however, which exact settings of the ICC test were used.

Longitudinal deformation is highest in the endotherefore

views. In the HUNT (Nord-Trøndelag Health) study,

where

LS

is

measured. So far, there is no sufficient evidence to decide if endocardial, midwall, or full wall strain is the best choice for clinical use. In this study, endocardial strain was analyzed for purely practical reasons because it was the only LS parameter that could be provided by all vendors.

In the present study, we found the averaged absolute difference between repeated measurements of different LS parameters ranging from 2.5% to 5.0% (irrespective of the 1 outlier of 6.4%, which can be attributed to timing issues). Although the lower end of this range might be still considered acceptable under certain conditions, the higher end constitutes an average relative error in the range of 25%, which renders a segmental strain measurement jeopardized

SEGMENTAL

for clinical use. Our analysis revealed that the

STRAIN. The feasibility of assessing segmental strain

segmental strain reproducibility can differ between

was different among vendors, which is likely due to

apical, mid, and basal segments, which again likely

both differences in image quality and tracking algo-

reflects

rithm. The differences observed among GE, Epsilon,

respective software packages.

and TOMTEC can be solely attributed to the applied

INTERVENDOR VARIABILITY. To our knowledge, this

software package because the same image datasets

is the first study to investigate the intervendor dif-

were used for analysis.

ferences of segmental strain in a clinical setting using

FEASIBILITY

OF

ASSESSING

We also found that the anterior and the lateral

different

underlying

algorithms

in

the

8 different software packages. The maximal absolute

walls have the highest rate of exclusions. This is in agreement with previous reports (10) and may be related to the high burden of artifacts and noise in this region leading to a poorer recognition of

T A B L E 5 Absolute Difference  SD of PS Measurement Among All Vendors

Hitachi

speckles.

Esaote

GE

Samsung

Siemens

VARIABILITY OF REPEATED STRAIN MEASUREMENTS

Esaote


3.4  6.8

GE


4.6  7.5
1.0  7.6

(TEST-RETEST VARIABILITY). A number of studies

Samsung
1.7  6.7

1.8  7.4 2.8  6.5

evaluated the intraobserver and interobserver vari-

Siemens


3.8  6.9

0.1  7.1 0.6  7.1

ability of segmental strain with conflicting results.

Toshiba


3.1  3.5

0.4  5.6

1.5  6.0
1.3  6.2 0.8  6.3

Mavinkurve-Groothuis et al. (17) assessed the repro-

Epsilon


3.3  7.2

0.1  7.3

1.2  5.7

ducibility of segmental strain in 1 vendor in a small

TOMTEC


1.9  7.3

1.5  7.3 2.4  6.8
0.5  7.6

number of normal volunteers and found that it was good in 4-chamber views but poor in the 2-chamber

Abbreviation as in Table 2.

Toshiba

Epsilon


2.0  7.0
1.6  6.4 0.3  7.2
0.5  5.9 1.8  7.2

1.0  6.4 1.3  7.0

8

Mirea et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017 - 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

F I G U R E 5 Segmental PS Strain Values of Each Vendor Versus the Strain Value of the Same Segment Averaged Over All Vendors

The y-axis represents PS strain values of each vendor; the x-axis represents the strain value of the same segment averaged over all vendors. The correlation coefficient (R2) and the slope of the regression line (a) are displayed in green. The line of identity is shown in red.

F I G U R E 6 Characteristics of Speckle Tracking Performance of the Different Vendors

difference between the vendor with the highest and the lowest values was 4.5% for all 3 measured strain parameters, which is slightly higher than the 3.7% earlier reported for GLS (3). The higher reproducibility of GLS could be related to the averaging algorithms over larger regions of the myocardium and inclusion of models of LV behavior. In contrast, segmental strain has to rely entirely on the accuracy of the local tracking results and the quality of the artifact detection algorithms. Several ANOVA post hoc tests showed significant differences among vendors (Figure 4). Different definitions of endocardial strain may partially explain these differences. Although some vendors calculate strain values at a virtual endocardial border, others track subendocardially (up to one-third of myocardial thickness), which might result in lower values. Lacking a reliable “ground truth” in this clinical

The x-axis shows the vendor-specific range of strain values measured within our popu-

setting, we compared segmental measurements of

lation. To reduce the impact of outliers, it was calculated as the average range of all

each vendor with the average of all vendors for the

vendors times the slope of the regression line from Figure 5. We interpret higher values as indicators of a higher fidelity in following segmental myocardial deformation. The y-axis

same segment (Figure 5). It is reassuring that we

shows the averaged absolute difference of segmental peak systolic strain measure-

found a moderate to good linear correlation for all

ments. High values indicate a high test-retest variability.

vendors, which is in contrast to earlier studies (18). Analysis per LV level revealed that the basal segments

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017

Mirea et al.

- 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

had the lowest agreement which is in agreement with

and scar extend or scar transmurality would reach far

previous studies (19).

beyond the scope of this reproducibility study.

Although the correlation coefficients indicated overall good intervendor agreement, the absolute segmental differences and limits of agreement were relatively

large,

indicating

that

a

considerable

amount of noise is superimposed on the overall bias in measurements between vendors (Table 5).

CONCLUSIONS In contrast to GLS measurements, which showed excellent reproducibility and only a moderate, yet significant bias between vendors (5), segmental LS measurements have a higher degree of measurement

GOOD LOCAL TRACKING VERSUS SUSCEPTIBILITY

variability. Therefore, single segmental strain mea-

FOR NOISE. If we assume from the above that the

surements should be used for clinical decision-

test-retest variability reflects to a large extent the

making, monitoring, and research only with caution.

robustness of the tracking algorithm and that a wide

The extent to which other means of segmental func-

range of values is an indicator of good fidelity to

tion assessment, such as strain curve shape analysis

segmental abnormalities, then a good software

or

should combine both. We have therefore combined

compensate for the relatively poor segmental repro-

both measures in a comprehensive graph (Figure 6).

ducibility remains to be determined.

relative

comparison

between

regions,

could

STUDY LIMITATIONS. Although the present study

was set up to mimic clinical routine, several parameters were controlled to prepare an optimal

APPENDIX

environment for a fair comparison of different machines and software packages: patients were

This paper is published on behalf of the EACVI-ASE-

selected for better than average image quality;

Industry Standardization Task Force chaired by

repeated scans were performed by the same expert

Luigi P. Badano, Padua (EACVI) and James D. Thomas,

examiner; except for Esaote (portable device), only

Chicago (ASE), and the participating companies,

high-end ultrasound machines were used; and a

represented by: Jamie Hamilton (Epsilon), Stefano

company representative ensured technically optimal

Pedri (Esaote), Peter Lysyansky (GE), Gunnar Hansen

acquisitions. Moreover, the analysis was performed

(GE), Yasuhiro Ito (Hitachi), Tomoaki Chono (Hitachi),

by an expert observer (with solid background in

Jane Vogel (Philips), David Prater (Philips), Sungwook

tissue Doppler and speckle tracking analysis). We

Park (Samsung), Jin Yong Lee (Samsung), Helene

must

of

Houle (Siemens), Bogdan Georgescu (Siemens), Rolf

segmental strain measurements will be even larger

Baumann (TOMTEC), Bernhard Mumm (TOMTEC),

in a real-world clinical setting. It must be further

Yashuhiko Abe (Toshiba), and Willem Gorissen

assumed that involving different software versions

(Toshiba).

therefore

assume

that

the

variability

of the same vendor would add to the disagreement of measurements. The interobserver variability was

ACKNOWLEDGMENTS The

not tested. It is, however, expected that it is even

industry partners for their active support and

higher.

constructive

contribution

authors to

this

thank project.

all The

In this study, we have tested the accuracy and

authors also thank Sarah Magits for her excellent

reproducibility of different peak strain parameters

logistic support and help with patient recruitment;

only. We did not investigate how reliably the shape of

our technicians Sarah Fabré, Ibn Tielens, Monique

a strain curve is reproduced by a software or how

Tillekaerts, Anita Tuteleers, and Jolien Vissers; our

well

are

colleagues, doctoral students, research fellows; and

reflected independent from their absolute strain

assistants Claire Bouleti, Guido Claessen, Charlien

values. Furthermore, we have not tested the repro-

Gabriels,

ducibility of the timing of strain peaks, which would

Thibault Petit, Frédéric Schnell, Daisy Thijs, and

be relevant for any dyssynchrony related function

Katrien De Vadder for their help with patient

assessment. All this remains a task for the further

scanning and data processing.

relative

differences

between

regions

Kaatje

Goetschalckx,

Peter

Haemers,

analysis of this dataset. For this paper, cardiac magnetic resonance was

ADDRESS FOR CORRESPONDENCE: Prof. Dr. Jens-

used solely to characterize patients with an infarct

Uwe Voigt, Department of Cardiovascular Diseases,

because it was our intention to include patients

University Hospital Gasthuisberg, Herestraat 49,

with a wide range of abnormalities. A detailed

3000 Leuven, Belgium. E-mail: jens-uwe.voigt@

investigation on the relation between strain values

uzleuven.be.

9

10

Mirea et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2017 - 2017:-–-

Variability of Segmental Longitudinal Strain Measurements

PERSPECTIVES COMPETENCY IN MEDICAL KNOWLEDGE: This

curve shape, are more robust and specific markers of

study deals with an essential topic: that of intervendor

disease remains to be determined.

reproducibility of segmental LS, and provides a comprehensive comparison between different software

TRANSLATIONAL OUTLOOK: GLS has demonstrated

packages. Our findings show that segmental strain

clinical relevance and is now implemented in daily prac-

measurements have a considerable test-retest and

tice. Segmental LS parameters could also provide a broad

intervendor bias, suggesting that such measurements

spectrum of diagnostic options; however, additional im-

should be used with prudence in research and clinical

provements from companies are still required to increase

practice. Whether other characteristics of strain, such as

the reproducibility of segmental strain measurements.

REFERENCES 1. Motoki H, Borowski AG, Shrestha K, et al. Incremental prognostic value of assessing left ventricular myocardial mechanics in patients with chronic systolic heart failure. J Am Coll Cardiol 2012;60:2074–81. 2. Lancellotti P, Cosyns B, Zacharakis D, et al. Importance of left ventricular longitudinal function and functional reserve in patients with degenerative mitral regurgitation: assessment by two-dimensional speckle tracking. J Am Soc Echocardiogr 2008;21:1331–6. 3. Farsalinos KE, Daraban AM, Ünlü S, Thomas JD, Badano LP, Voigt JU. Head-to-head comparison of global longitudinal strain measurements among nine different vendors: the EACVI/ASE InterVendor Comparison Study. J Am Soc Echocardiogr 2015;28:1171–81. 4. Lang RM, Badano LP, Mor-Avi V, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2015;16:233–70. 5. Plana JC, Galderisi M, Barac A, et al. Expert consensus for multimodality imaging evaluation of adult patients during and after cancer therapy: a report from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2014;15:1063–9. 6. Phelan D, Collier P, Thavendiranathan P, et al. Relative apical sparing of longitudinal strain using two-dimensional speckle-tracking echocardiography is both sensitive and specific for the diagnosis of cardiac amyloidosis. Heart 2012;98:1442–8. 7. Bertini M, Ng AC, Antoni ML, et al. Global longitudinal strain predicts long-term survival in

patients with chronic ischemic cardiomyopathy. Circ Cardiovasc Imaging 2012;5:383–91.

transmural extent of healed Q-wave and non-Qwave myocardial infarction. Lancet 2001;357:21–8.

8. Voigt JU, Arnold MF, Karlsson M, et al. Assessment of regional longitudinal myocardial strain rate derived from Doppler myocardial imaging indexes in normal and infarcted myocardium. J Am Soc Echocardiogr 2000;13:588–98.

15. Mada RO, Lysyansky P, Daraban AM, Duchenne J, Voigt JU. How to define end-diastole and end-systole? Impact of timing on strain measurements. J Am Coll Cardiol Img 2015;8:148–57.

9. Reant P, Zaroui A, Donal E, et al. Identification and characterization of super-responders after cardiac resynchronization therapy. Am J Cardiol 2010;105:1327–35.

16. Leitman M, Lysiansky M, Lysyansky P, et al. Circumferential and longitudinal strain in 3 myocardial layers in normal subjects and in patients with regional left ventricular dysfunction. J Am Soc Echocardiogr 2010;23:64–70.

10. Barbier P, Mirea O, Cefalù C, Maltagliati A, Savioli G, Guglielmo M. Reliability and feasibility of longitudinal AFI global and segmental strain compared with 2D left ventricular volumes and ejection fraction: intra- and inter-operator, testretest, and inter-cycle reproducibility. Eur Heart J

17. Mavinkurve-Groothuis AM, Weijers G, GrootLoonen J, et al. Interobserver, intraobserver and intrapatient reliability scores of myocardial strain imaging with 2-D echocardiography in patients treated with anthracyclines. Ultrasound Med Biol 2009;35:697–704.

Cardiovasc Imaging 2015;16:642–52.

18. Takigiku K, Takeuchi M, Izumi C, et al., JUSTICE investigators. Normal range of left ventricular 2-

11. Thorstensen A, Dalen H, Amundsen BH, Aase SA, Stoylen A. Reproducibility in echocardiographic assessment of the left ventricular global and regional function, the HUNT study. Eur J Echocardiogr 2010;11:149–56. 12. Thomas JD, Badano LP. EACVI-ASE-industry initiative to standardize deformation imaging: a brief update from the co-chairs. Eur Heart J

dimensional strain: Japanese Ultrasound Speckle Tracking of the Left Ventricle (JUSTICE) study. Circ J 2012;76:2623–32. 19. Castel AL, Szymanski C, Delelis F, et al. Prospective comparison of speckle tracking longitudinal bidimensional strain between two vendors. Arch Cardiovasc Dis 2014;107:96–104.

Cardiovasc Imaging 2013;14:1039–40. 13. Voigt JU, Pedrizzetti G, Lysyansky P, et al. Definitions for a common standard for 2D speckle tracking echocardiography: consensus document of the EACVI/ASE/Industry Task Force to standardize deformation imaging. Eur Heart J Cardiovasc Imaging 2015;16:1–11. 14. Wu E, Judd RM, Vargas JD, Klocke FJ, Bonow RO, Kim RJ. Visualisation of presence, location, and

KEY WORDS intervendor bias, segmental strain

A PPE NDI X For supplemental material, figures, and tables, please see the online version of this article.