Shear Wave elastography on the Toshiba Aplio 500

Transcription

1 A comparison of one shot and continuous modes for Shear Wave elastography on the Toshiba Aplio 500 Sandra O Hara 1,2, MMS DMU AMS AFASA 1 SKG Radiology, West Perth, Perth, Western Australia 2 Department of Medical Radiation Sciences, Curtin University, Perth, Western Australia, Australia

2

3 Sandra O Hara 1,2, MMS DMU AMS AFASA 1 SKG Radiology, West Perth, Perth, Western Australia 2 Department of Medical Radiation Sciences, Curtin University, Perth, Western Australia, Australia Sandra O Hara is the Tutor Sonographer at SKG Radiology in Perth, overseeing one of the largest Ultrasound training programs in Australia. With over 25 years of experience as a Sonographer, Sandra obtained the General DMU in She is experienced in general, obstetric, vascular and musculoskeletal ultrasound, and is committed to maintaining high standards of practice in the field of Sonography. Sandra completed a Master of Medical Sonography in 2015 with the research for this qualification incorporating 3 peer reviewed publications. More recently Sandra has been awarded WA Sonographer of the year and has become an Associate Fellow of the ASA. Introduction When using the shear wave elastography mode, the Toshiba Aplio 500 system supplies two methods to obtain readings of meters per second (m/s), and kilopascals (kpa) to assess the stiffness of tissues in the human body; the one-shot and continuous methods. The one-shot method activates the main pulse over one frame to measure the resultant speed of the shear wave and elasticity in the tissue being examined. A single image is produced to register the stiffness measurements. The continuous mode provides a live acquisition of the elastographic and propagation maps, and multiple images are acquired over time. It is also possible to cine-loop back through the frames to register multiple stiffness measurements from the one acquisition. The main pulse used in shear wave technology is a sound beam that is modified to a high intensity in order to produce the shear wave propagation.the one-shot method is considered to be the most accurate method clinically as it provides a truer representation of data at a particular point in time, and at a particular location. The one-shot method also provides a higher amplitude push pulse, and a greater number of tracking pulses. Following the application of the one-shot method, a mandatory crystal cool time of 5 seconds is usually observed before a further tissue interrogation can be made. In continuous mode, multiple frames can be sampled and stored in a cine loop with 3 levels of frame rates available. The frame rate level of 1 equates to 0.4 frames per second (fps); with levels 2 and 3 providing 0.7 and 1 fps respectively, with small delays during image acquisition to allow for crystal cooling. The compounding of images in the continuous mode can result in greater errors in measurements as higher levels of noise may be introduced into the output. Higher frame rates produce weaker amplitudes of the main pulse and a reduced number of tracking pulses. The ideal clinical situation is to use the continuous mode to find the most appropriate anatomical window for shear wave interrogation and then activate the one-shot method to obtain data. Some patients however, may be noncompliant to breathing instructions and technically it may be difficult to achieve results using a one-shot interrogation. This investigation therefore compares the use and measurements of the one-shot and continuous modes on the adult liver using the Aplio 500 system.

4 Figure 1: Speed and propagation maps displaying 2.5cm x 3cm elastogram and two 10mm ROI s registering the mean speed and SD for the ROI. Method Data was collected over a 3-month period at SKG Radiology in Perth, Western Australia. All data was collected using the Toshiba Aplio 500 series (Japan) (Version 6 & 6.5) paired with the 6C1 curvilinear transducer. The mean age of the 30 participants was 40 years (range 24 57), with 18 of the participants having no history of liver disease and 12 with a confirmed diagnosis of Hepatitis C. All participants had their livers interrogated using both continuous and one-shot modes. The continuous mode measurements were registered at the lowest frame rate setting of 1, or 0.4 fps. Continuous mode measurements were registered with gentle respiration, and one-shot mode with gentle cessation of respiration during expiration. The same imaging methodology was utilised for both one-shot and continuous modes. As demonstrated in figure 1, an elastogram box size of 2.5cm x 3cm with a 10mm circular ROI was used to obtain the mean speed and elasticity of the shear wave. A maximum of two ROIs were placed within each speed or pressure map and concordant propagation map simultaneously, at similar depths within the liver. All ROI s were obtained in anatomical segments 7 or 8 of the liver. All of the mean measurements were registered with an output standard deviation of less than 20%. The mean output from the region of interest (ROI) was registered for each participant a minimum of 13 times. Outlying measurements were excluded so as to obtain a maximally homogenous dataset of 10 measurements. In Transient elastography (TE) based techniques, measurements of tissue elasticity registered in kpa are obtained, and 10 measurements are used to assess for homogeneity. This is achieved by assessing the interquartile measurement for the dataset. The interquartile range is divided by the median for the data set and multiplied by 100 to calculate a percentage. A percentage of less than 30% is considered to be a homogenous dataset. Currently in TE and shear wave elastography of the liver the median of the 10 registered means is the reported measurement. Reflecting the standard for the reporting of liver fibrosis, the median value of the 10 measurements was used to compare the one shot and continuous modes. To assess for homogeneity of the shear wave dataset of 10 means registered in both one shot and continuous modes, the registered interquartile range was divided by the median for all participants as described previously. All datasets achieved a median/interquartile range of less than 30% for the registered kpa values with the same calculation for the registered speed or m/s values equating to approximately half of the kpa value. This calculation is used to mirror results obtained from the use of Transient elastography (TE) in the assessment of stiffness of liver tissue.

5 Statistical analysis Data analysis was performed using SPSS version 26.0 (SPSS V26.0, Chicago, USA). Descriptive data were presented as mean ± standard deviation (SD). The variables were input to assess normality using a Kolmogorov-Smirnov Test. The data did not differ significantly (p>0.05) from normality with exclusion of the one case outlying from the normal distribution (29 participants). The authentic outlier at higher median speed was included in the analysis with no effect on the analysis observed. The one-shot method was used as the reference measurement, and the continuous method was assessed for difference from the one-shot method. We defined any observed difference between the two measurements as measurement bias. Bias was assessed in two ways. (1) The mean of the array of 30 differences, one difference from each participant, was compared to an ideal value of zero using a 1-sample t-test. The null hypothesis, H0: mean bias = 0, was tested against a two-sided alternative, at the 5% level of statistical significance, (p<0.05). (2) Additionally, the bias values were plotted against the averages of the pairs of measured lengths using Bland-Altman plots (Bland & Altman, 1986) This provided a visual indication of how any measurement bias varies, if at all, with a range of input shear wave speed and elasticity values in respect of both magnitude and trend. Results One-shot and continuous mode measurements were obtained in all 30 participants. For the one-shot mode and continuous modes respectively, the mean of the median shear wave speed values was 1.46 m/s (SD 0.40) and 1.48 m/s (SD 0.41.) respectively with mean values of 6.87 kpa (SD 5.58) and kpa (SD 5.8) for the respective elasticity measurements. The mean difference for the speed of tissue for continuous mode and one shot mode was not statistically different from zero with a mean difference of m/s (SE 0.16); t(30)=-0.99,p =0.33. The Bland Altman plot seen at figure 2 shows a large number of values near the zero bias line, and a very slight positive bias of 0.016m/s. The registered oneshot and continuous mode median speed outputs obtained for each of the 30 cases is also presented at figure 3 in the format of a scatterplot, which also displays the line of equal speed. The mean difference for the elasticity of tissue for continuous mode and one-shot mode was not statistically different from zero, with a mean difference of kpa (SE 0.15); t(30)=-0.95,p =0.35. The Bland Altman plot seen at figure 4 shows a large number of values near the zero bias line and a very slight positive bias of 0.143kPa. The registered one-shot and continuous mode median kpa outputs obtained for each of the 30 cases is also presented at figure 5 in the format of a scatterplot, which also displays the line of equal elasticity. Figure 2. Figure 3.

6 Figure 4. Figure 5. Discussion The one-shot shear wave acquisition mode is recognised as ideal clinically, as a single frame is acquired with increased penetration and high resolution. The continuous mode acquires multiple frames with decreased penetration and increased noise, which may compound errors arising from image noise. The continuous mode is however recognised as advantageous in assisting the operator to identify the best anatomical window to obtain reliable shear wave propagation. Comparative results demonstrate that the Aplio 500 system delivers highly correlated results between one-shot and continuous modes. We therefore contend that during the usertraining phase, the continuous mode utilised at its lowest frame rate, may be considered an acceptable method for the novice operator to obtain reliable shear wave measurements. As operator skill increases however, the one-shot mode should become the standard method used clinically. Additionally, we have shown that in a non-compliant patient the continuous mode, at its lowest frame rate, is likely an appropriate method with which to obtain clinically relevant results on this system. The Toshiba Aplio 500 platinum ultrasound system supplies a shear wave elastography application that is unique from other vendors. The operator can assess the reliability of the shear wave propagation through multiple maps that provide a pre-emptive indication of the reliability of placement of the ROI before an analysis. The elastogram maps are presented to show both speed of arrival time of the shear wave and elasticity (pressure) of tissues. Loss of colour filling in the speed and pressure map is indicative of regions where shear wave propagation has not been successful. The propagation map is unique to Toshiba. This map shows the user areas where increased reliability is more likely. Regions registering a parallel arrangement of propagation lines have increased reliability for resultant shear wave speed. Regions where there is a loss of parallel configuration show decreased reliability and thus should be avoided. The propagation map is also a visual aid to determine tissue stiffness; the closer the lines are the softer the tissue, the wider apart they are the stiffer the tissue. The Toshiba system also supplies the standard deviation of the registered mean speed or elasticity measurement, which is useful in the context of several areas of tissue being examined. It is recommended by the vendor that the registered SD of the mean should be kept to less than 20% of the mean, or an artefactual increase in the mean speed may be registered in the results. All registered measurements in this research achieved a standard deviation of less than 20%. These measures of reliability checking assist the operator in achieving a placement of the ROI in regions where a reliable shear wave propagation output is expected. Figure 6 is an example of the speed map indicative of shear wave speed utilised in dual mode; with the propagation map showing the

7 ROIs placed in the region of greatest reliability, and as is demonstrated a low SD of the mean has been achieved. Figure 7 is an example of ROI placement in a technically challenging patient. This research shows that it is possible with this system to achieve an extremely close replication of results comparing the one-shot mode with continuous mode; as long as the frame rate is kept at the minimum level for the continuous mode method. The results show a statistically insignificant difference between the median speed and elasticity measurements for one shot and continuous modes. As can be seen in figures 2 and 4, one value was registered above the 95% confidence interval and some values approached this interval, thus we need to consider if these differences will be medically significant. An overestimation of shear wave values may alter the patient treatment regime and the tools provided by the Toshiba Aplio 500 system to quantify the reliability of the shear wave propagation should be considered when deciding which mode is most appropriate for the examination. Figure 6: Shear wave propagation and ROI placement. Figure 7: Shear wave propagation and ROI placement demonstrating a technically challenging case with one reliable ROI placement achieved. Conclusion The Toshiba Aplio Shear Wave application supplies the user with multiple indicators of the reliability of the shear wave propagation. These tools should be used in all cases. The operator should consider using a combination of continuous and one-shot modes, especially when working with a non-compliant patient. The reliability of the propagation map will be synchronous with the registered standard deviation of the mean, and therefore these maps and SD assessments should always be utilized. Regions of higher reliability will register a lower SD and a reading registered with an SD greater than 20% of the mean should be considered unreliable. The one-shot mode is recognised as the ideal method to obtain reliable data of shear wave speed and elasticity of tissues. Continuous mode, even though shown in our study to give very similar results to one-shot mode in most patients, should only be used if one-shot mode is unsuccessful; and always at the lowest frame rate available. In conclusion the mode that achieves the best propagation maps and lower standard deviations should be the preferred mode for the shear wave examination. Reference Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing agreement between two methods of clinical measurement., The Lancet, February 8.

8