Skip to main content
. 2024 Aug 1;8:86. doi: 10.1186/s41747-024-00483-9

Table 2.

Reproducibility of fully automated vBMDL1-4 measurements in each group

Group name Group size Measurement 1 vBMDL1-4 [mg CaHA/cm3] Measurement 2 vBMDL1-4 [mg CaHA/cm3] Absolute difference [mg CaHA/cm3] Relative difference [%] RMSE [mg CaHA/cm3] RMSCV [%]
Number Mean ± standard deviation Mean ± standard deviation Mean ± standard deviation Mean ± standard deviation
Δrecona 169 175.3 ± 54.8 175.4 ± 54.7 0.1 ± 3.0 0.2 ± 2.1 3.0 1.3
Δcontrasta 133 171.2 ± 53.7 170.1 ± 55.3 -1.0 ± 6.4 -0.9 ± 4.5 6.1 2.9
Δsessiona 123 160.1 ± 54.0 162.8 ± 58.9 2.7 ± 12.5 1.3 ± 7.4 11.0 4.7
Δsession_no-outliersa 111 156.4 ± 51.4 157.0 ± 53.6 0.6 ± 9.5 0.3 ± 5.9 9.0 4.1
Δscannera 63 164.1 ± 56.6 162.0 ± 58.3 -2.1 ± 8.9 -2.0 ± 7.3 8.7 4.9
Δalla 20 139.2 ± 43.7 131.5 ± 45.9 -7.7 ± 8.0 -6.3 ± 6.7 7.6 5.9
Scanner-agnostic 171 190.8 ± 51.7 191.1 ± 49.3 0.4 ± 10.1 0.9 ± 5.9 10.1 3.7

a Group contains asynchronously calibrated measurements only. vBMD Volumetric bone mineral density, CaHA Calcium hydroxylapatite, RMSE Root mean square error, RMSCV Root mean square coefficient of variation. Δsession_no-outliers comprises the Δsession group following exclusion of outliers, which exceeded the inner boundaries of the 95% confidence interval limit of agreement of the Bland Altman plot in Fig. 4