Reliability Criteria for 2D-SWE of the Liver: IQR/Median and Number of Samples
Number of Measurements for 2D-SWE
According to EFSUMB recommendations (Dietrich et al., 2017, Recommendation 13; LoE 2b, GoR B, strong consensus 18/0/0), a minimum of three measurements should be obtained for 2D-SWE, and the result should be expressed as the median along with the interquartile range (IQR). WFUMB (Ferraioli et al., 2024) confirms that a minimum of three individual samples is sufficient for calculating a reliable LSM for 2D-SWE; no differences in diagnostic accuracy were found between averaging five and three samples when liver biopsy is used as a reference.
ACR/SRU (2024) indicates that for 2D-SWE, at least five measurements may be appropriate when using a quality assessment parameter. In comparison, for pSWE, it is recommended to obtain 10 measurements from 10 independent images at one location and use the median (ACR/SRU 2024; WFUMB 2018/2024).
Quality Indicator IQR/M
The ratio of the interquartile range to the median (IQR/M) is used as a measure of quality. According to ACR/SRU (2024):
| Units of Measurement | IQR/M Threshold for Acceptable Data Set |
|---|---|
| kPa | ≤30% |
| m/s | ≤15% |
An IQR/M value of ≤30% (for kPa) is generally accepted as an indicator of an accurate data set (ACR/SRU 2024); for pSWE, an IQR/M <30% (in kPa) improves the accuracy of fibrosis staging (WFUMB 2018, EFSUMB 2017).
Placement of ROI and Depth
The region of interest (ROI) should be placed away from large vessels, bile ducts, and mass lesions (ACR/SRU 2024). The operator places the ROI in an area free of large vessels, at a depth greater than 2 cm but not deeper than 8 cm (EFSUMB Course Book, Sporea et al., 2018). Measurement accuracy decreases at depths greater than 6 cm from the transducer due to ARFI pulse attenuation (WFUMB 2018).
Note: The provided excerpts lack specific IQR/M criteria and the number of samples specifically marked as “EFSUMB 2024”. [clarify]
Frequently asked questions
How many measurements are minimally required for 2D-SWE of the liver?
A minimum of three valid measurements (EFSUMB 2017; WFUMB 2024). ACR/SRU (2024) considers at least five appropriate when using a quality parameter.
How should the result of 2D-SWE be expressed?
As the median along with the interquartile range (IQR) — the median and IQR are robust to non-normal data distribution (EFSUMB 2017).
What is the acceptable IQR/M threshold?
IQR/M ≤30% for values in kPa and ≤15% for values in m/s (ACR/SRU 2024).
Does accuracy differ when averaging three versus five samples?
No differences in diagnostic accuracy were found between averaging five and three samples when liver biopsy is used as a reference (WFUMB 2024).
How should the ROI be placed?
Away from large vessels, bile ducts, and lesions, at a depth >2 cm and not deeper than 8 cm; accuracy decreases deeper than 6 cm due to ARFI attenuation (EFSUMB Course Book 2018; WFUMB 2018).