Article Figures & Data
Figures
- Fig 1.
Labeling plane on the coronary (A) and transverse (B) MRA images. Arrows show RICA (R), LICA (L), and vertebrobasilar arteries (B).
- Fig 2.
For the 6 patients (18 vessels) with 3 PLDs, there are no significant differences in the supply fractions with relative labeling efficiencies (P > .05).
- Fig 3.
ROC curves of VE-PCASL with standard (blue line) and relative (green line) labeling efficiencies, respectively, for the identification of feeding arteries, and the AUC is larger for relative labeling efficiencies.
- Fig 4.
Patient 1, a 23-year-old man. He had headaches and blurred vision in both eyes for half a year. The supply fractions of the left ICA, VA, and right ICA were 52.59%, 46.75%, 0.66% and 49.27%, 50.71%, 0 with standard and relative labeling efficiencies, respectively, showing good agreement between the 2 methods. A, VE-PCASL with standard labeling efficiency (arrows show the AVM lesion). B, DSA shows that the AVM lesion is fed by the left ICA and VA.
- Fig 5.
Patient 13, a 27-year-old woman. She had a sudden severe headache accompanying a visual field defect 2 months ago, which was confirmed as spontaneous intraventricular hemorrhage by CT. A, VE-PCASL with standard labeling efficiencies shows relatively poor separation of left-versus-right ICA perfusion territories, and there are abnormally high signals in the left ICA perfusion territory with the supply fraction of 14.31%, mismatching the results of DSA. B, VE-PCASL with relative labeling efficiencies shows an improved separation, and there are fewer abnormally high signals from the left ICA perfusion territory with the supply fraction of 1.14% (arrows show the AVM lesion). C, DSA shows that the AVM lesion is fed by the VA and right ICA.
- Fig 6.
Patient 9, a 31-year-old woman. She had dizziness for approximately 2 years. VE-PCASL with standard labeling efficiencies showed abnormally high signals in the left ICA perfusion territory with the supply fraction of 27.53%, mismatching the results of DSA. A, VE-PCASL with relative labeling efficiencies shows reduced abnormal signals with the supply fraction of 25.67% but still fails to match the results of DSA (arrows show the AVM lesion). B, DSA shows that the AVM lesion is fed by the VA and right ICA.
Tables
- Table 1:
Modified Hadamard encoding matrix based on estimated relative labeling efficiencies of the 18 patients with AVMs
A −1.0000 −1.0000 −1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 −0.9200 ± 0.083 0.1600 ± 0.186 0.7100 ± 0.340 1.0000 0.9200 ± 0.083 −0.1600 ± 0.186 −0.7100 ± 0.340 1.0000 0.9700 ± 0.105 −0.5300 ± 0.112 0.9700 ± 0.105 1.0000 −0.9700 ± 0.105 0.5300 ± 0.112 −0.9700 ± 0.105 1.0000 A+ −0.1087 ± 0.0312 0.1087 ± 0.0312 −0.3067 ± 0.0210 0.3067 ± 0.0210 0.1125 ± 0.0387 −0.1125 ± 0.0387 −0.3233 ± 0.031 0.3233 ± 0.031 0.0000 −0.0000 −0.3333 ± 0.0402 0.3333 ± 0.0402 −0.0680 ± 0.0446 0.0680 ± 0.0446 0.3067 ± 0.0210 −0.3067 ± 0.0210 0.2209 ± 0.0528 −0.2209 ± 0.0528 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 Note:—Data are means.
- Table 2:
Demographic data on the studied population and characterization information of AVMs
Patient No. Age (yr) Sex Spetzler-Martin Classification Major Feeding Arteries Diagnosed by DSA RICA LICA VA 1 23 M S2E1V0 0 1 1 2 21 M S2E0V1 1 1 1 3 34 M S2E0V0 0 1 1 4 44 M S2E1V1 1 1 1 5 27 F S3E1V0 1 0 1 6 22 F S3E1V1 1 1 1 7a 40 M S2E0V0 0 1 1 8a 46 F S2E0V0 1 1 1 9a 31 F S2E1V0 1 0 1 10a 47 F S2E1V0 1 0 1 11a 13 M S1E0V0 0 1 1 12a 39 M S2E0V0 1 0 0 13 27 F S2E1V0 1 0 1 14 31 M S3E1V0 1 1 1 15 25 M S1E1V0 0 1 1 16 47 M S2E0V0 1 0 1 17 39 F S1E1V1 0 1 1 18 28 F S1E0V0 0 1 0 Note:—S indicates the size of the cerebral AVM nidus (1, <3 cm; 2, 3∼6 cm; 3, >6 cm); E, eloquence of adjacent brain (0, noneloquent; 1, eloquent); V, venous drainage pattern (0, superficial draining vein; 1, deep draining vein).
↵a Patients who had 3 PLD time points. Zero and 1 indicate arteries that have or have not contributed to AVM lesions.
- Table 3:
Different cutoff values of the supply fraction of feeding arteries of the AVM and the resulting sensitivity and specificity with standard or relative labeling efficiencies respectively
Cutoff Value Standard Relative Sensitivity Specificity Sensitivity Specificity 0% 100% 0 100% 6.67% 5% 97.44% 40% 97.44% 66.67% 10% 92.31% 73.33% 89.74% 80% 15% 84.62% 93.33% 82.05% 93.33% 20% 76.92% 93.33% 79.49% 93.33% 25% 74.36% 93.33% 69.23% 93.33%
