Multicenter Test-Retest and Acquisition Protocol Reproducibility of Liver Iron Quantification

Author

Lu Mao

Statistical analysis

For test-retest repeatability of R2*, the Bland-Altman analysis was used to calculate the bias, limits of agreement (LoA), and coefficient of repeatability (half-width of LoA) by field strength, protocol, liver segment, lobe, and overall. Between-protocol reproducibility was assessed by similar Bland-Altman analysis by field strength, liver segment, lobe, and overall. Per guidelines of Raunig et al. (2014), the intra-class correlation coefficient (ICC) was used for repeatability analysis and the concordance correlation coefficient (CCC) for reproducibility analysis. P-values \(<\) 0.05 were considered statistically significant. All analyses were performed in R version 4.3.3 (R Foundation for Statistical Computing, Vienna, Austria).

Patient characteristics

Patient characteristics are summarized by site in Table 1.

Table 1: Patient characteristics by site.
Characteristic UW, N = 591 UTSW, N = 481 JHU, N = 521 Stanford, N = 461 Overall, N = 2051
Age (years) 51 (28, 59) 46 (34, 56) 30 (20, 53) 17 (12, 22) 36 (19, 54)
    Unknown 0 0 0 1 1
Sex




    f 20 (34%) 23 (48%) 29 (56%) 18 (40%) 90 (44%)
    m 39 (66%) 25 (52%) 23 (44%) 27 (60%) 114 (56%)
    Unknown 0 0 0 1 1
Weight (kg) 74 (66, 96) 75 (61, 102) 65 (55, 79) 53 (35, 67) 68 (56, 87)
    Unknown 0 1 0 6 7
Height (cm) 173 (163, 179) 172 (161, 182) 165 (159, 173) 157 (141, 167) 168 (158, 178)
    Unknown 0 1 0 6 7
Ferriscan LIC (mg/g) 4 (2, 8) 2 (1, 5) 6 (2, 15) 3 (1, 8) 3 (1, 8)
    Unknown 2 0 2 3 7
Ferriscan R2 (/s) 74 (51, 111) 43 (33, 83) 98 (50, 163) 65 (41, 117) 66 (40, 116)
    Unknown 2 0 2 3 7
Serum Ferritin (ng/mL) 659 (363, 1,702) 850 (466, 1,748) 1,231 (390, 3,095) 822 (577, 1,399) 889 (440, 1,918)
    Unknown 0 0 1 3 4
Disease Categories




    1 22 (37%) 20 (42%) 17 (33%) 7 (16%) 66 (33%)
    2 13 (22%) 20 (42%) 32 (63%) 28 (64%) 93 (46%)
    3 0 (0%) 1 (2.1%) 0 (0%) 2 (4.5%) 3 (1.5%)
    4 15 (25%) 1 (2.1%) 1 (2.0%) 6 (14%) 23 (11%)
    5 7 (12%) 0 (0%) 0 (0%) 0 (0%) 7 (3.5%)
    6 2 (3.4%) 6 (13%) 1 (2.0%) 1 (2.3%) 10 (5.0%)
    Unknown 0 0 1 2 3
1 Median (IQR); n (%)

Repeatability analysis

The number of pairs for test-retest repeatability analysis by protocol, field strength, and lobe is listed below.

Number of test-retest pairs in analysis
Protocol Field Lobe n
P01 1.5T Left lobe 480
P01 1.5T Right lobe 384
P01 3.0T Left lobe 470
P01 3.0T Right lobe 376
P02 1.5T Left lobe 480
P02 1.5T Right lobe 384
P02 3.0T Left lobe 460
P02 3.0T Right lobe 368
P04 1.5T Left lobe 475
P04 1.5T Right lobe 380
P04 3.0T Left lobe 430
P04 3.0T Right lobe 344
P08 1.5T Left lobe 455
P08 1.5T Right lobe 364
P08 3.0T Left lobe 425
P08 3.0T Right lobe 340

Overall

The test-retest bias, coefficient of repeatability (CR), intra-class correlation (ICC) coefficient and 95% confidence interval (CI) are summarized by protocol and field strength across all segments/lobes in Table 2 below.

Table 2: Test-retest repeatability analysis by protocol and field strength across all segments/lobes.
Protocol Field Bias Bias (%) (LoA %) CR ICC (95% CI) p-value
P01 1.5T 0.9 0.5% (-21.1%, 22.2%) 54.9 (0.217) 0.994 (0.993, 0.995) < 0.001
P01 3.0T -6.4 -0.2% (-18.5%, 18.1%) 129.7 (0.183) 0.991 (0.99, 0.992) < 0.001
P02 1.5T 1.3 0.2% (-19.9%, 20.2%) 53.2 (0.201) 0.994 (0.993, 0.994) < 0.001
P02 3.0T 2.8 0.4% (-19.9%, 20.7%) 135.5 (0.203) 0.989 (0.987, 0.99) < 0.001
P04 1.5T 0.2 0.5% (-21.6%, 22.7%) 70 (0.222) 0.988 (0.986, 0.989) < 0.001
P04 3.0T 1.4 0.3% (-32%, 32.6%) 155.5 (0.323) 0.975 (0.971, 0.978) < 0.001
P08 1.5T 0.8 0.2% (-25.2%, 25.7%) 66.1 (0.255) 0.987 (0.985, 0.989) < 0.001
P08 3.0T 2.3 0.9% (-30.7%, 32.5%) 134.9 (0.316) 0.965 (0.96, 0.97) < 0.001

Bland-Altman plots of test-retest repeatability by protocol for 1.5T and 3.0T are shown in Figure 1 and Figure 2, respectively.

Figure 1: Bland-Altman plot of test-retest repeatability of R2* in 1.5T by protocol.
Figure 2: Bland-Altman plot of test-retest repeatability of R2* in 3.0T by protocol.

By lobe

The test-retest bias, coefficient of repeatability (CR), intra-class correlation (ICC) coefficient and 95% confidence interval (CI) are summarized by protocol, field strength, and lobe in Table 3 below.

Table 3: Test-retest repeatability analysis by protocol, field strength, and lobe.
Protocol Field Lobe Bias Bias (%) (LoA %) CR ICC (95% CI) p-value
P01 1.5T Left 2.2 0.6% (-23.6%, 24.7%) 58.1 (0.242) 0.993 (0.992, 0.994) < 0.001
P01 1.5T Right -0.7 0.5% (-17.6%, 18.5%) 50.5 (0.181) 0.995 (0.994, 0.996) < 0.001
P01 3.0T Left -8.6 -0.3% (-21%, 20.4%) 147 (0.207) 0.988 (0.985, 0.99) < 0.001
P01 3.0T Right -3.7 -0.1% (-15%, 14.8%) 104.1 (0.149) 0.995 (0.994, 0.996) < 0.001
P02 1.5T Left 1.0 0.2% (-21.8%, 22.1%) 57.8 (0.22) 0.992 (0.99, 0.993) < 0.001
P02 1.5T Right 1.7 0.1% (-17.4%, 17.6%) 46.8 (0.175) 0.996 (0.995, 0.996) < 0.001
P02 3.0T Left 7.6 0.8% (-22.8%, 24.3%) 159.4 (0.236) 0.982 (0.978, 0.985) < 0.001
P02 3.0T Right -3.3 -0.1% (-15.4%, 15.2%) 96.7 (0.153) 0.995 (0.994, 0.996) < 0.001
P04 1.5T Left 0.3 0.5% (-25.1%, 26%) 71.5 (0.255) 0.986 (0.983, 0.988) < 0.001
P04 1.5T Right 0.1 0.6% (-16.5%, 17.6%) 68.1 (0.171) 0.99 (0.988, 0.992) < 0.001
P04 3.0T Left 0.7 0.4% (-32.9%, 33.8%) 152.5 (0.334) 0.968 (0.961, 0.973) < 0.001
P04 3.0T Right 2.2 0.1% (-30.9%, 31.1%) 159.5 (0.31) 0.98 (0.975, 0.983) < 0.001
P08 1.5T Left 1.4 0.5% (-26.4%, 27.4%) 66.6 (0.269) 0.985 (0.982, 0.987) < 0.001
P08 1.5T Right 0.1 -0.1% (-23.6%, 23.4%) 65.5 (0.235) 0.989 (0.986, 0.991) < 0.001
P08 3.0T Left 1.2 1.1% (-34.9%, 37.1%) 124.7 (0.36) 0.956 (0.948, 0.964) < 0.001
P08 3.0T Right 3.7 0.7% (-24.5%, 25.8%) 146.8 (0.252) 0.97 (0.963, 0.976) < 0.001

Bland-Altman plots of test-retest repeatability by protocol and lobe for 1.5T and 3.0T are shown in Figure 3 and Figure 4, respectively.

Figure 3: Bland-Altman plot of test-retest repeatability of R2* in 1.5T by protocol and lobe.
Figure 4: Bland-Altman plot of test-retest repeatability of R2* in 3.0T by protocol and lobe.

Reproducibility analysis

The number of observations for between-protocol reproducibility analysis by field strength and lobe is listed below.

Number of observations in between-protocol reproducibility analysis
Protocol Field Lobe n
P01 1.5T Left lobe 990
P01 1.5T Right lobe 792
P01 3.0T Left lobe 960
P01 3.0T Right lobe 768
P02 1.5T Left lobe 975
P02 1.5T Right lobe 780
P02 3.0T Left lobe 925
P02 3.0T Right lobe 740
P03 1.5T Left lobe 490
P03 1.5T Right lobe 392
P03 3.0T Left lobe 455
P03 3.0T Right lobe 364
P04 1.5T Left lobe 975
P04 1.5T Right lobe 780
P04 3.0T Left lobe 905
P04 3.0T Right lobe 724
P05 1.5T Left lobe 480
P05 1.5T Right lobe 384
P05 3.0T Left lobe 450
P05 3.0T Right lobe 360
P06 1.5T Left lobe 455
P06 1.5T Right lobe 364
P06 3.0T Left lobe 440
P06 3.0T Right lobe 352
P07 1.5T Left lobe 485
P07 1.5T Right lobe 388
P07 3.0T Left lobe 440
P07 3.0T Right lobe 352
P08 1.5T Left lobe 945
P08 1.5T Right lobe 756
P08 3.0T Left lobe 875
P08 3.0T Right lobe 700
P09 1.5T Left lobe 435
P09 1.5T Right lobe 348
P09 3.0T Left lobe 360
P09 3.0T Right lobe 288
P10 1.5T Left lobe 415
P10 1.5T Right lobe 332
P10 3.0T Left lobe 375
P10 3.0T Right lobe 300

Overall

The between-protocol bias, coefficient of reproducibility (CR), concordance correlation coefficient (CCC) and 95% confidence interval (CI) are summarized across all segments/lobes for 1.5T and 3.0 T in Table 4 and Table 5, respectively.

CCC heatmap

Pairwise CCC is visualized in Figure 5 below.

Figure 5: Pairwise CCC for between-protocol reproducibility.

1.5T

Table 4: Between-protocol reproducibility analysis across all segments/lobes for 1.5T.
Protocol x Protocol y Bias Bias (%) (LoA %) CR CCC (95% CI)
P01 P02 12.8 5.5% (-17.8%, 28.8%) 74.9 (0.233) 0.99 (0.989, 0.99)
P03 35.3 7.2% (-32.1%, 46.5%) 282.8 (0.393) 0.86 (0.845, 0.874)
P04 20.6 5.3% (-24.8%, 35.5%) 178.5 (0.302) 0.938 (0.933, 0.943)
P05 51.5 13.1% (-25.1%, 51.3%) 302.1 (0.382) 0.825 (0.807, 0.841)
P06 34.5 10.5% (-22.6%, 43.6%) 176.2 (0.331) 0.915 (0.906, 0.923)
P07 59.6 14.5% (-29.3%, 58.3%) 353 (0.438) 0.752 (0.729, 0.773)
P08 39.1 10.7% (-24.2%, 45.7%) 240.2 (0.349) 0.876 (0.867, 0.885)
P09 60.6 12.6% (-42.1%, 67.3%) 411.2 (0.547) 0.684 (0.653, 0.714)
P10 68.5 14.6% (-45.5%, 74.8%) 448.4 (0.601) 0.654 (0.619, 0.687)
P02 P03 21.4 1.1% (-40.7%, 42.9%) 251.3 (0.418) 0.885 (0.871, 0.897)
P04 7.7 -0.4% (-30.4%, 29.6%) 152.2 (0.3) 0.954 (0.949, 0.957)
P05 38.5 7.1% (-33.8%, 48%) 262.5 (0.409) 0.859 (0.845, 0.873)
P06 21.9 3.9% (-32.4%, 40.2%) 145.6 (0.363) 0.94 (0.933, 0.946)
P07 47.4 8.8% (-37.5%, 55.1%) 316.7 (0.463) 0.786 (0.765, 0.805)
P08 26.7 5.2% (-30.3%, 40.6%) 210.8 (0.354) 0.9 (0.892, 0.907)
P09 47.1 6.5% (-51%, 63.9%) 375.4 (0.574) 0.719 (0.689, 0.747)
P10 52.6 8.2% (-53.7%, 70%) 410.3 (0.619) 0.686 (0.652, 0.717)
P03 P04 -7.1 -0.7% (-32.3%, 31%) 138.2 (0.316) 0.961 (0.956, 0.966)
P05 17.7 6% (-26.1%, 38.2%) 142.7 (0.322) 0.95 (0.944, 0.956)
P06 19.9 5% (-28.2%, 38.3%) 138.1 (0.333) 0.945 (0.938, 0.951)
P07 27.2 7.8% (-31.2%, 46.8%) 188.3 (0.39) 0.906 (0.894, 0.916)
P08 17.0 6.1% (-27.9%, 40.1%) 144.8 (0.34) 0.949 (0.943, 0.955)
P09 26.7 6.5% (-36.9%, 49.9%) 188.6 (0.434) 0.909 (0.897, 0.92)
P10 29.1 7.4% (-42%, 56.9%) 218 (0.494) 0.888 (0.873, 0.901)
P04 P05 24.7 6.9% (-21.4%, 35.1%) 134.6 (0.283) 0.955 (0.95, 0.96)
P06 22.2 5.3% (-27.2%, 37.8%) 142.8 (0.325) 0.941 (0.934, 0.948)
P07 34.2 8.6% (-27.7%, 45%) 195.5 (0.364) 0.901 (0.89, 0.911)
P08 20.0 5.9% (-23.3%, 35.1%) 132.9 (0.292) 0.954 (0.95, 0.958)
P09 34.1 6.7% (-42.1%, 55.6%) 240.4 (0.489) 0.861 (0.844, 0.877)
P10 37.2 8.1% (-44.9%, 61.1%) 271.4 (0.53) 0.836 (0.816, 0.855)
P05 P06 4.1 -0.5% (-30.6%, 29.6%) 112 (0.301) 0.962 (0.957, 0.967)
P07 10.0 2.1% (-34.1%, 38.2%) 158.4 (0.362) 0.93 (0.92, 0.938)
P08 -0.5 -0.1% (-33.7%, 33.4%) 127.2 (0.336) 0.959 (0.953, 0.964)
P09 8.9 0.3% (-45.7%, 46.4%) 188.4 (0.46) 0.905 (0.891, 0.917)
P10 11.9 1.4% (-46.9%, 49.7%) 208.9 (0.483) 0.892 (0.876, 0.906)
P06 P07 1.8 1.8% (-32.8%, 36.4%) 125.5 (0.346) 0.949 (0.942, 0.956)
P08 -6.0 0.4% (-31%, 31.7%) 115.7 (0.313) 0.961 (0.955, 0.965)
P09 -2.7 -0.8% (-38.3%, 36.7%) 117 (0.375) 0.959 (0.953, 0.965)
P10 -0.6 0.4% (-42.2%, 43%) 153.3 (0.426) 0.935 (0.925, 0.944)
P07 P08 -10.3 -2% (-39.5%, 35.4%) 153.1 (0.374) 0.935 (0.926, 0.943)
P09 -0.5 -1.6% (-44.6%, 41.5%) 152.7 (0.43) 0.933 (0.923, 0.942)
P10 1.0 -0.2% (-48%, 47.6%) 180.9 (0.478) 0.913 (0.9, 0.924)
P08 P09 7.7 0.1% (-44.9%, 45.1%) 172.7 (0.45) 0.92 (0.909, 0.93)
P10 12.7 1.6% (-47.5%, 50.6%) 206.7 (0.491) 0.895 (0.88, 0.909)
P09 P10 -1.1 0.8% (-40.9%, 42.6%) 133.6 (0.417) 0.949 (0.942, 0.956)

3.0T

Table 5: Between-protocol reproducibility analysis across all segments/lobes for 3.0T.
Protocol x Protocol y Bias Bias (%) (LoA %) CR CCC (95% CI)
P01 P02 31.4 5.9% (-18.3%, 30.2%) 241.4 (0.243) 0.968 (0.965, 0.97)
P03 114.3 11.4% (-57.9%, 80.6%) 721.6 (0.693) 0.65 (0.616, 0.681)
P04 94.9 9.9% (-52.1%, 71.9%) 643.8 (0.62) 0.707 (0.687, 0.727)
P05 164.7 19.6% (-58.8%, 98.1%) 842 (0.785) 0.495 (0.456, 0.531)
P06 173.0 20.5% (-59.3%, 100.3%) 859.8 (0.798) 0.467 (0.428, 0.504)
P07 156.1 20.4% (-54.4%, 95.2%) 752.7 (0.748) 0.552 (0.515, 0.586)
P08 140.1 17.1% (-49.8%, 84%) 754.4 (0.669) 0.551 (0.526, 0.575)
P09 50.5 3.8% (-63.5%, 71.1%) 501.4 (0.673) 0.746 (0.714, 0.775)
P10 177.8 20.3% (-72.3%, 113%) 934.3 (0.926) 0.442 (0.395, 0.488)
P02 P03 69.4 4.5% (-57.9%, 67%) 638.1 (0.625) 0.685 (0.651, 0.715)
P04 54.5 3.1% (-52.4%, 58.6%) 544.6 (0.555) 0.766 (0.748, 0.783)
P05 109.4 12% (-56.5%, 80.5%) 689 (0.685) 0.59 (0.554, 0.625)
P06 125.8 14.4% (-61.1%, 89.9%) 723.7 (0.755) 0.541 (0.502, 0.578)
P07 103.9 12.6% (-52.1%, 77.4%) 641 (0.648) 0.624 (0.588, 0.657)
P08 105.3 11.2% (-54.3%, 76.7%) 671 (0.655) 0.595 (0.57, 0.618)
P09 8.0 -2.8% (-61.9%, 56.2%) 361.1 (0.591) 0.845 (0.822, 0.865)
P10 120.7 13.4% (-74.9%, 101.6%) 796.6 (0.883) 0.518 (0.47, 0.563)
P03 P04 5.0 1.2% (-43.7%, 46.1%) 238.5 (0.449) 0.936 (0.927, 0.944)
P05 44.6 8.6% (-36.3%, 53.5%) 280.3 (0.449) 0.895 (0.881, 0.907)
P06 61.3 11.3% (-45.2%, 67.8%) 331.8 (0.565) 0.847 (0.828, 0.864)
P07 54.7 11% (-41.7%, 63.8%) 314.7 (0.528) 0.864 (0.847, 0.88)
P08 38.9 8% (-32.9%, 48.8%) 253 (0.409) 0.912 (0.9, 0.922)
P09 13.8 -0.5% (-62.7%, 61.7%) 362.2 (0.622) 0.84 (0.817, 0.861)
P10 46.0 9.8% (-61.8%, 81.4%) 385 (0.716) 0.832 (0.808, 0.853)
P04 P05 40.1 8% (-40.3%, 56.2%) 237.9 (0.483) 0.918 (0.908, 0.928)
P06 54.0 10% (-42.4%, 62.5%) 269.5 (0.524) 0.89 (0.876, 0.902)
P07 47.6 9.7% (-39.3%, 58.8%) 232.8 (0.491) 0.918 (0.908, 0.928)
P08 43.5 7.6% (-34.3%, 49.6%) 268.7 (0.419) 0.895 (0.886, 0.903)
P09 5.5 -2.2% (-62.7%, 58.4%) 302 (0.605) 0.883 (0.865, 0.899)
P10 34.3 8% (-61.2%, 77.1%) 298 (0.691) 0.892 (0.876, 0.906)
P05 P06 15.7 2.8% (-51%, 56.6%) 239.3 (0.538) 0.91 (0.897, 0.921)
P07 9.0 2.3% (-44.3%, 48.8%) 224.8 (0.465) 0.923 (0.912, 0.932)
P08 -2.1 0.1% (-41.9%, 42.1%) 199.5 (0.42) 0.939 (0.93, 0.947)
P09 -26.7 -8.2% (-70.1%, 53.7%) 316.8 (0.619) 0.853 (0.83, 0.873)
P10 -3.4 0.7% (-64.2%, 65.5%) 323.2 (0.649) 0.863 (0.842, 0.882)
P06 P07 -6.6 -0.6% (-54.8%, 53.6%) 247.1 (0.542) 0.904 (0.89, 0.916)
P08 -17.8 -2.7% (-54.2%, 48.8%) 216.5 (0.515) 0.924 (0.913, 0.934)
P09 -43.4 -10.9% (-74.1%, 52.4%) 306.1 (0.632) 0.85 (0.827, 0.87)
P10 -19.1 -1.5% (-71.6%, 68.7%) 303.4 (0.701) 0.874 (0.855, 0.891)
P07 P08 -15.7 -2.9% (-47.7%, 41.8%) 216.8 (0.448) 0.924 (0.913, 0.934)
P09 -38.3 -10.6% (-69.3%, 48.1%) 286 (0.587) 0.871 (0.851, 0.889)
P10 -17.2 -2.6% (-65.3%, 60.1%) 277.9 (0.627) 0.897 (0.881, 0.911)
P08 P09 -24.2 -8.5% (-66.7%, 49.8%) 296.1 (0.583) 0.871 (0.851, 0.888)
P10 6.1 2.3% (-68.5%, 73.1%) 293.7 (0.708) 0.878 (0.859, 0.895)
P09 P10 17.4 5.8% (-49.9%, 61.5%) 254.7 (0.557) 0.908 (0.893, 0.921)

Bland-Altman plots

Bland-Altman plots of between-protocol reproducibility for 1.5T and 3.0T are shown in Figure 6 and Figure 7, respectively.

Figure 6: Bland-Altman plot of between-protocol reproducibility of R2* in 1.5T.
Figure 7: Bland-Altman plot of between-protocol reproducibility of R2* in 3.0T.

By lobe

The between-protocol bias, coefficient of reproducibility (CR), concordance correlation coefficient (CCC) and 95% confidence interval (CI) are summarized by lobe for 1.5T and 3.0 T in Table 6 and Table 7, respectively.

CCC heatmap

Pairwise CCC is visualized in Figure 8 below.

Figure 8: Pairwise CCC for between-protocol reproducibility by lobe.

1.5T

Table 6: Between-protocol reproducibility analysis of R2* by lobe in 1.5T.
Lobe Protocol x Protocol y Bias Bias (%) (LoA %) CR CCC (95% CI)
Left P01 P02 13.3 5.6% (-19.1%, 30.3%) 87.2 (0.247) 0.985 (0.983, 0.987)
P03 37.4 8.2% (-34.2%, 50.6%) 295.3 (0.424) 0.833 (0.808, 0.855)
P04 22.7 6% (-27.5%, 39.5%) 191.4 (0.335) 0.923 (0.914, 0.931)
P05 54.2 14.4% (-27.9%, 56.7%) 311.2 (0.423) 0.794 (0.767, 0.819)
P06 37.5 11.9% (-24.1%, 48%) 195.7 (0.36) 0.89 (0.874, 0.904)
P07 56.8 14.9% (-28.8%, 58.7%) 327.5 (0.437) 0.77 (0.74, 0.797)
P08 40.5 11.4% (-26.5%, 49.4%) 239.6 (0.38) 0.866 (0.853, 0.878)
P09 59.7 13% (-44.4%, 70.4%) 393.8 (0.574) 0.687 (0.643, 0.726)
P10 68.0 15.3% (-45.4%, 76%) 429.5 (0.607) 0.655 (0.607, 0.698)
P02 P03 23.5 2.1% (-41.9%, 46%) 254.7 (0.44) 0.867 (0.846, 0.886)
P04 9.6 0.3% (-33.6%, 34.2%) 162.4 (0.339) 0.941 (0.934, 0.948)
P05 41.0 8.4% (-35.3%, 52%) 262.8 (0.436) 0.841 (0.818, 0.861)
P06 25.2 5.3% (-32.5%, 43.1%) 156.4 (0.378) 0.925 (0.914, 0.935)
P07 44.4 9.2% (-36.9%, 55.3%) 283.1 (0.461) 0.813 (0.787, 0.836)
P08 27.7 6% (-32.4%, 44.3%) 205.7 (0.384) 0.894 (0.883, 0.905)
P09 44.4 6.3% (-53.7%, 66.2%) 348.3 (0.6) 0.735 (0.695, 0.77)
P10 52.1 8.5% (-54.3%, 71.3%) 386.2 (0.628) 0.694 (0.649, 0.735)
P03 P04 -8.2 -0.9% (-36.2%, 34.3%) 153.2 (0.353) 0.947 (0.937, 0.955)
P05 18.6 6.4% (-29.1%, 41.9%) 143.5 (0.355) 0.942 (0.932, 0.951)
P06 22.1 5.9% (-31.2%, 43%) 154.4 (0.371) 0.925 (0.912, 0.936)
P07 22.1 7.2% (-32.7%, 47.2%) 170.4 (0.399) 0.917 (0.903, 0.929)
P08 16.1 6% (-29.7%, 41.8%) 144.5 (0.357) 0.944 (0.934, 0.952)
P09 22.8 5.7% (-38.5%, 49.8%) 162 (0.441) 0.927 (0.913, 0.938)
P10 25.9 7% (-38.8%, 52.8%) 182.5 (0.458) 0.913 (0.897, 0.927)
P04 P05 26.4 7.4% (-23.4%, 38.3%) 142.7 (0.309) 0.944 (0.935, 0.951)
P06 25.3 6.3% (-28.8%, 41.5%) 161.1 (0.351) 0.919 (0.906, 0.931)
P07 30.1 8.2% (-28.4%, 44.8%) 170.6 (0.366) 0.919 (0.906, 0.93)
P08 20.5 6.1% (-26.8%, 39.1%) 146 (0.329) 0.939 (0.932, 0.945)
P09 31.6 6.1% (-46%, 58.3%) 230.9 (0.521) 0.862 (0.838, 0.882)
P10 35.8 7.8% (-46.1%, 61.6%) 258.9 (0.539) 0.839 (0.811, 0.863)
P05 P06 5.2 0.1% (-33.9%, 34.1%) 122.3 (0.34) 0.949 (0.939, 0.958)
P07 3.9 0.9% (-36%, 37.9%) 139.2 (0.37) 0.941 (0.93, 0.951)
P08 -2.5 -0.7% (-39.7%, 38.2%) 144.2 (0.39) 0.94 (0.929, 0.95)
P09 3.1 -1.3% (-49.8%, 47.1%) 172.9 (0.485) 0.911 (0.892, 0.926)
P10 7.1 0.3% (-49.3%, 49.8%) 198.6 (0.495) 0.892 (0.87, 0.91)
P06 P07 -1.9 0.8% (-37.6%, 39.2%) 134.7 (0.384) 0.936 (0.924, 0.947)
P08 -7.4 -0.3% (-34.9%, 34.2%) 124.3 (0.346) 0.949 (0.94, 0.958)
P09 -3.3 -1.9% (-41.9%, 38.1%) 125.5 (0.4) 0.948 (0.937, 0.957)
P10 -4.1 -0.9% (-41.7%, 39.9%) 146.2 (0.408) 0.936 (0.922, 0.947)
P07 P08 -6.0 -1.4% (-39.6%, 36.8%) 140.5 (0.382) 0.942 (0.931, 0.951)
P09 0.3 -1.9% (-50.8%, 47%) 170.1 (0.489) 0.913 (0.895, 0.928)
P10 2.9 0% (-49.9%, 49.8%) 178.2 (0.499) 0.911 (0.892, 0.926)
P08 P09 3.7 -1% (-50.1%, 48.1%) 173.3 (0.491) 0.913 (0.895, 0.927)
P10 10.5 1.1% (-50.3%, 52.5%) 197.2 (0.514) 0.897 (0.876, 0.914)
P09 P10 -0.5 1% (-43.8%, 45.8%) 130.9 (0.448) 0.948 (0.937, 0.957)
Right P01 P02 12.3 5.4% (-15.9%, 26.8%) 55.8 (0.213) 0.994 (0.994, 0.995)
P03 32.7 5.9% (-28.9%, 40.7%) 266.8 (0.348) 0.888 (0.87, 0.905)
P04 18.0 4.5% (-20.7%, 29.8%) 160.9 (0.252) 0.954 (0.949, 0.959)
P05 48.1 11.4% (-20.6%, 43.5%) 290.8 (0.32) 0.855 (0.834, 0.874)
P06 30.8 8.7% (-20.1%, 37.4%) 148.1 (0.287) 0.943 (0.934, 0.95)
P07 63.1 14% (-29.9%, 58%) 382.9 (0.439) 0.733 (0.696, 0.766)
P08 37.3 9.9% (-20.8%, 40.6%) 241 (0.307) 0.887 (0.874, 0.898)
P09 61.8 12.2% (-39%, 63.3%) 432.5 (0.511) 0.682 (0.634, 0.724)
P10 69.1 13.8% (-45.7%, 73.2%) 471.6 (0.595) 0.654 (0.6, 0.701)
P02 P03 18.6 -0.2% (-39%, 38.7%) 247.2 (0.389) 0.902 (0.884, 0.917)
P04 5.4 -1.3% (-25.6%, 23%) 138.4 (0.243) 0.966 (0.961, 0.97)
P05 35.4 5.5% (-31.6%, 42.5%) 262.4 (0.37) 0.878 (0.858, 0.895)
P06 17.7 2.1% (-31.9%, 36.2%) 130.6 (0.341) 0.956 (0.948, 0.962)
P07 51.1 8.3% (-38.3%, 54.9%) 354.4 (0.466) 0.758 (0.723, 0.789)
P08 25.5 4.2% (-27.2%, 35.5%) 217.1 (0.313) 0.906 (0.894, 0.916)
P09 50.5 6.8% (-47.4%, 60.9%) 407.1 (0.542) 0.703 (0.656, 0.744)
P10 53.3 7.8% (-52.9%, 68.5%) 439.1 (0.607) 0.678 (0.625, 0.724)
P03 P04 -5.6 -0.3% (-26.8%, 26.2%) 116.8 (0.265) 0.975 (0.97, 0.98)
P05 16.6 5.6% (-21.9%, 33%) 142 (0.275) 0.957 (0.949, 0.965)
P06 17.3 4% (-23.6%, 31.5%) 114.5 (0.276) 0.965 (0.959, 0.971)
P07 33.7 8.6% (-29.2%, 46.4%) 208.1 (0.378) 0.894 (0.875, 0.91)
P08 18.1 6.1% (-25.6%, 37.8%) 145.4 (0.317) 0.954 (0.945, 0.962)
P09 31.7 7.6% (-35%, 50.1%) 217.1 (0.425) 0.891 (0.869, 0.909)
P10 33.1 7.9% (-45.8%, 61.7%) 255.5 (0.537) 0.862 (0.834, 0.886)
P04 P05 22.5 6.1% (-18.5%, 30.8%) 123.7 (0.246) 0.967 (0.961, 0.972)
P06 18.3 4% (-24.7%, 32.6%) 115.8 (0.287) 0.965 (0.958, 0.97)
P07 39.2 9.1% (-26.9%, 45.2%) 222.6 (0.361) 0.882 (0.863, 0.9)
P08 19.5 5.6% (-18.1%, 29.3%) 114.6 (0.237) 0.969 (0.965, 0.972)
P09 37.3 7.5% (-37%, 52%) 251.9 (0.445) 0.861 (0.835, 0.883)
P10 38.9 8.5% (-43.4%, 60.4%) 286.6 (0.519) 0.834 (0.801, 0.861)
P05 P06 2.7 -1.3% (-25.5%, 23%) 97.9 (0.243) 0.975 (0.969, 0.979)
P07 17.6 3.5% (-31.5%, 38.5%) 178.7 (0.35) 0.918 (0.902, 0.932)
P08 2.1 0.6% (-24.6%, 25.8%) 102 (0.252) 0.977 (0.972, 0.981)
P09 16.3 2.5% (-40%, 44.9%) 205.6 (0.425) 0.899 (0.877, 0.917)
P10 18.0 2.8% (-43.9%, 49.5%) 220.9 (0.467) 0.891 (0.867, 0.911)
P06 P07 6.5 3% (-26%, 32.1%) 112.5 (0.29) 0.963 (0.954, 0.969)
P08 -4.3 1.2% (-25.5%, 27.9%) 104 (0.267) 0.972 (0.965, 0.977)
P09 -1.9 0.6% (-33.3%, 34.6%) 105.6 (0.34) 0.97 (0.963, 0.976)
P10 3.8 2% (-42.6%, 46.6%) 161.6 (0.446) 0.935 (0.919, 0.948)
P07 P08 -15.7 -2.8% (-39.3%, 33.7%) 167.1 (0.365) 0.928 (0.914, 0.94)
P09 -1.6 -1.1% (-35.3%, 33.2%) 128.1 (0.343) 0.956 (0.946, 0.964)
P10 -1.3 -0.4% (-45.6%, 44.8%) 184.3 (0.452) 0.915 (0.896, 0.932)
P08 P09 12.7 1.5% (-37.6%, 40.6%) 171.7 (0.391) 0.928 (0.912, 0.941)
P10 15.5 2.2% (-43.8%, 48.3%) 218.2 (0.461) 0.893 (0.87, 0.913)
P09 P10 -1.9 0.6% (-37.1%, 38.4%) 137 (0.377) 0.951 (0.939, 0.96)

3.0T

Table 7: Between-protocol reproducibility analysis of R2* by lobe in 3.0T.
Lobe Protocol x Protocol y Bias Bias (%) (LoA %) CR CCC (95% CI)
Left P01 P02 33.1 6.4% (-21.5%, 34.3%) 280.6 (0.279) 0.952 (0.946, 0.958)
P03 131.9 14.3% (-60.3%, 88.9%) 757.5 (0.746) 0.574 (0.524, 0.62)
P04 110.9 12.5% (-55.6%, 80.7%) 678.6 (0.681) 0.626 (0.593, 0.656)
P05 175.4 22% (-61.5%, 105.4%) 836.1 (0.834) 0.433 (0.38, 0.483)
P06 184.4 23.5% (-61.8%, 108.8%) 852 (0.853) 0.405 (0.352, 0.456)
P07 174.5 23.3% (-56.9%, 103.5%) 797 (0.802) 0.449 (0.398, 0.498)
P08 151.8 19.6% (-53.4%, 92.6%) 765.3 (0.73) 0.475 (0.44, 0.508)
P09 62.3 4.7% (-70%, 79.4%) 569.8 (0.747) 0.661 (0.606, 0.709)
P10 188.2 21.9% (-73.8%, 117.5%) 912.7 (0.956) 0.395 (0.332, 0.454)
P02 P03 79.2 6.6% (-60.6%, 73.7%) 646.3 (0.671) 0.635 (0.586, 0.68)
P04 67.0 5.2% (-55.2%, 65.5%) 562.5 (0.603) 0.708 (0.679, 0.734)
P05 117.1 13.6% (-60.1%, 87.3%) 681 (0.737) 0.544 (0.492, 0.592)
P06 135.7 16.9% (-63.8%, 97.5%) 725.9 (0.807) 0.473 (0.417, 0.525)
P07 118.0 14.9% (-55.5%, 85.3%) 670.3 (0.704) 0.535 (0.483, 0.583)
P08 114.2 13.1% (-58.7%, 84.9%) 676.9 (0.718) 0.523 (0.488, 0.557)
P09 14.6 -2.8% (-68.5%, 62.9%) 414.2 (0.657) 0.783 (0.742, 0.819)
P10 124.2 13.9% (-74.9%, 102.7%) 749.5 (0.888) 0.496 (0.432, 0.556)
P03 P04 7.3 1.3% (-49.2%, 51.8%) 267.7 (0.505) 0.9 (0.881, 0.916)
P05 42.3 8.2% (-41%, 57.3%) 272.6 (0.492) 0.874 (0.852, 0.893)
P06 58.9 11.8% (-50.3%, 73.9%) 320.7 (0.621) 0.819 (0.789, 0.845)
P07 55.0 11.3% (-45.5%, 68%) 320.4 (0.567) 0.818 (0.788, 0.844)
P08 35.0 7.8% (-36.1%, 51.8%) 221.3 (0.44) 0.917 (0.902, 0.929)
P09 5.5 -2.6% (-69.2%, 64%) 350.5 (0.666) 0.824 (0.787, 0.854)
P10 37.8 7.8% (-64.6%, 80.2%) 341.8 (0.724) 0.83 (0.796, 0.859)
P04 P05 36.1 7.4% (-44.1%, 58.8%) 240.8 (0.515) 0.897 (0.878, 0.913)
P06 50.7 10.4% (-43.7%, 64.4%) 259.4 (0.54) 0.872 (0.85, 0.891)
P07 47.5 10% (-41.2%, 61.3%) 245.8 (0.512) 0.884 (0.864, 0.901)
P08 38.9 7.4% (-37.4%, 52.2%) 265.2 (0.448) 0.87 (0.855, 0.883)
P09 -5.9 -4.3% (-66.1%, 57.4%) 284 (0.618) 0.875 (0.848, 0.897)
P10 25.9 6.1% (-64.1%, 76.3%) 293.6 (0.702) 0.868 (0.841, 0.891)
P05 P06 15.5 3.5% (-54.9%, 62%) 237 (0.584) 0.887 (0.865, 0.905)
P07 11.1 2.7% (-47.6%, 53%) 222.2 (0.503) 0.9 (0.881, 0.916)
P08 -4.3 0.3% (-46.7%, 47.2%) 200.1 (0.47) 0.925 (0.91, 0.937)
P09 -35.5 -10.4% (-76.4%, 55.6%) 315.3 (0.66) 0.823 (0.787, 0.854)
P10 -8.2 -1.6% (-63.9%, 60.8%) 297.1 (0.623) 0.851 (0.819, 0.877)
P06 P07 -3.7 -0.6% (-57%, 55.8%) 222.7 (0.564) 0.897 (0.877, 0.915)
P08 -19.9 -3.4% (-58.5%, 51.7%) 198 (0.551) 0.92 (0.905, 0.934)
P09 -52.5 -13.4% (-79%, 52.1%) 311.5 (0.655) 0.808 (0.771, 0.841)
P10 -25.0 -4.5% (-73%, 63.9%) 285.7 (0.684) 0.852 (0.821, 0.878)
P07 P08 -19.7 -3.6% (-51.5%, 44.2%) 209.1 (0.478) 0.911 (0.893, 0.925)
P09 -48.9 -13.2% (-72.9%, 46.5%) 259.9 (0.597) 0.864 (0.837, 0.887)
P10 -26.3 -5.4% (-66.3%, 55.5%) 265.1 (0.609) 0.871 (0.845, 0.894)
P08 P09 -30.0 -10.5% (-72.1%, 51%) 284.2 (0.616) 0.858 (0.828, 0.883)
P10 -2.3 -0.6% (-71.3%, 70%) 252.9 (0.707) 0.891 (0.867, 0.91)
P09 P10 18.2 6.2% (-55.1%, 67.4%) 273.5 (0.613) 0.879 (0.852, 0.901)
Right P01 P02 29.4 5.3% (-13.4%, 24.1%) 180.7 (0.188) 0.983 (0.981, 0.985)
P03 92.4 7.7% (-53.6%, 69.1%) 672.6 (0.614) 0.729 (0.685, 0.768)
P04 75.0 6.6% (-46.1%, 59.3%) 595.6 (0.527) 0.786 (0.761, 0.808)
P05 151.4 16.7% (-54.7%, 88.2%) 849.9 (0.715) 0.554 (0.499, 0.606)
P06 158.7 16.8% (-55.1%, 88.6%) 869.9 (0.718) 0.527 (0.47, 0.58)
P07 133.2 16.7% (-50.2%, 83.5%) 692 (0.669) 0.659 (0.611, 0.703)
P08 125.4 14% (-43.9%, 71.9%) 740.1 (0.579) 0.625 (0.591, 0.657)
P09 35.9 2.7% (-54.1%, 59.5%) 399.1 (0.568) 0.847 (0.814, 0.874)
P10 164.9 18.4% (-70.4%, 107.1%) 961.7 (0.888) 0.488 (0.416, 0.555)
P02 P03 57.1 1.9% (-53.8%, 57.7%) 627.7 (0.557) 0.733 (0.687, 0.773)
P04 38.9 0.5% (-47.8%, 48.8%) 520.1 (0.483) 0.82 (0.797, 0.84)
P05 99.7 9.9% (-51.4%, 71.3%) 699.5 (0.613) 0.635 (0.581, 0.682)
P06 113.5 11.3% (-56.9%, 79.5%) 721.2 (0.682) 0.607 (0.552, 0.657)
P07 86.3 9.9% (-46.7%, 66.5%) 601.5 (0.566) 0.711 (0.665, 0.752)
P08 94.1 8.7% (-47.6%, 65.1%) 663.4 (0.563) 0.662 (0.628, 0.693)
P09 -0.2 -2.9% (-52.6%, 46.7%) 281 (0.496) 0.913 (0.892, 0.93)
P10 116.3 12.7% (-75%, 100.4%) 853 (0.877) 0.536 (0.461, 0.604)
P03 P04 2.2 1.1% (-35.7%, 37.9%) 196.2 (0.368) 0.965 (0.957, 0.972)
P05 47.3 9.1% (-29.8%, 48%) 289.9 (0.389) 0.91 (0.892, 0.925)
P06 64.1 10.6% (-38%, 59.3%) 345.6 (0.486) 0.867 (0.842, 0.889)
P07 54.4 10.7% (-36.7%, 58.2%) 308 (0.474) 0.898 (0.877, 0.915)
P08 43.7 8.2% (-28.4%, 44.8%) 287.8 (0.366) 0.907 (0.889, 0.922)
P09 24.1 2.1% (-53.8%, 58.1%) 375.9 (0.56) 0.854 (0.821, 0.882)
P10 56.4 12.2% (-58.2%, 82.7%) 432.8 (0.704) 0.832 (0.794, 0.863)
P04 P05 45.1 8.7% (-35.3%, 52.7%) 234.3 (0.44) 0.935 (0.921, 0.946)
P06 58.0 9.6% (-40.8%, 60%) 281.8 (0.504) 0.903 (0.884, 0.919)
P07 47.7 9.4% (-36.9%, 55.7%) 215.9 (0.463) 0.943 (0.932, 0.953)
P08 49.2 7.9% (-30.1%, 46%) 272.9 (0.381) 0.913 (0.902, 0.923)
P09 19.8 0.5% (-58.1%, 59.1%) 321.4 (0.586) 0.889 (0.863, 0.91)
P10 44.7 10.4% (-57.3%, 78%) 302.6 (0.676) 0.909 (0.888, 0.927)
P05 P06 15.9 1.9% (-45.6%, 49.4%) 242.4 (0.475) 0.926 (0.91, 0.94)
P07 6.4 1.7% (-39.7%, 43.1%) 228.2 (0.414) 0.938 (0.924, 0.95)
P08 0.6 -0.1% (-34.9%, 34.6%) 199 (0.348) 0.95 (0.939, 0.96)
P09 -15.6 -5.4% (-61.5%, 50.6%) 318 (0.56) 0.878 (0.848, 0.903)
P10 2.6 3.4% (-64.2%, 71.1%) 353.4 (0.676) 0.871 (0.84, 0.897)
P06 P07 -10.3 -0.6% (-52.1%, 51%) 274.8 (0.515) 0.908 (0.887, 0.925)
P08 -15.0 -1.9% (-48.6%, 44.7%) 237.8 (0.466) 0.926 (0.909, 0.94)
P09 -32.0 -7.6% (-67.4%, 52.1%) 298.2 (0.598) 0.885 (0.857, 0.908)
P10 -11.7 2.4% (-69.3%, 74%) 324.2 (0.716) 0.889 (0.862, 0.911)
P07 P08 -10.8 -2.1% (-42.7%, 38.5%) 226 (0.406) 0.934 (0.919, 0.946)
P09 -25.1 -7.4% (-64.3%, 49.6%) 314.1 (0.57) 0.876 (0.846, 0.901)
P10 -5.9 1% (-63.4%, 65.3%) 292 (0.644) 0.914 (0.892, 0.931)
P08 P09 -16.9 -5.9% (-59.5%, 47.7%) 310.2 (0.536) 0.881 (0.852, 0.905)
P10 16.7 6.1% (-64.3%, 76.5%) 337 (0.704) 0.866 (0.834, 0.892)
P09 P10 16.5 5.3% (-42.6%, 53.3%) 229.6 (0.48) 0.936 (0.919, 0.949)

References

Raunig, David L, Lisa M McShane, Gene Pennello, Constantine Gatsonis, Paul L Carson, James T Voyvodic, Richard L Wahl, et al. 2014. “Quantitative Imaging Biomarkers: A Review of Statistical Methods for Technical Performance Assessment.” Statistical Methods in Medical Research 24 (1): 27–67. https://doi.org/10.1177/0962280214537344.