Article Figures & Data
Figures
- FIG 1.
Diffusion data-processing pipeline for glioma assessment. Initial multishell DWI (A) is followed by Marchenko-Pastur principal component analysis (MP-PCA) denoising (B) to reduce noise. Subsequent steps include Gibbs ringing artifact removal (C), susceptibility-induced distortion (SID) correction (D), and N4 bias field correction (E) to improve image quality. Structural imaging (F) is used for tumor segmentation (G), which is then coregistered (H) with the diffusion data. Modeling of the diffusion data (J) enables extraction of tumor parameters (K), such as ADC, MK, and μFA. The I letter was intentionally omitted to improve readability.
- FIG 2.
Illustration of tumor segmentation and generation of tumoral masks. The figure demonstrates the segmentation of gliomas using 4 image sequences (from left to right and top to bottom: contrast-enhanced T1-weighted, T1-weighted, FLAIR, and T2-weighted). These masks are subsequently coregistered with diffusion data sets.
- FIG 3.
ROC curves for selected parameters illustrating the diagnostic accuracy in determining IDH status. The curves compare DKI (A), SMT parameters (B), and the standard clinically used ADC method (C), with an emphasis on parameters with AUC values of >0.90.
- FIG 4.
ROC curves for selected parameters illustrating the diagnostic accuracy in determining glioma grade.The curves compare DKI (A), and SMT (B), parameters with the standard clinically used ADC method (C), with an emphasis on parameters with AUC values of 0.90.
Tables
Characteristic Result No. of patients 80 Age, mean (yr) 48 (SD, 16) HGG 43 (54) IDHmt 46 (58) Male sex 49 (61) Postcontrast enhancement 46 (58) Necrosis 31 (39) T2 FLAIR mismatch 6 (8) Hemorrhage 24 (30) Note:—HGG indicates high-grade glioma; IDHmt, IDH-mutant.
↵a Unless otherwise noted, data represent the number of patients; data in parentheses are percentages.
- Table 2:
Comparative diagnostic accuracy of investigated parameters differentiating IDH-mutant from IDH-wild-type gliomasa
Models and Parameters AUC 95% CI Optimal Cutoff Sensitivity (%) Specificity (%) P Value Nondiffusion Enhancementb 0.77 0.66–0.88 0.5 ↓ 88.2 65.2 <.001 Necrosisb 0.85 0.77–0.94 0.5 ↓ 79.4 91.3 <.001 T2LFMb 0.57 0.44–0.70 0.5 ↑ 13.0 100 .03 Hemorrhageb 0.78 0.67–0.88 0.5 ↓ 61.8 93.5 <.001 ADC ADCc 0.82 0.73–0.92 1.206 ↑ 67.4 91.2 <.001 DTI FA 0.59 0.45–0.73 0.174 ↓ 50.0 82.6 .050 MDc 0.83 0.75–0.92 0.945 ↑ 89.1 64.7 <.001 DKI AK 0.90 0.83–0.97 0.583 ↓ 88.2 84.8 <.001 RK 0.90 0.83–0.97 0.593 ↓ 94.1 78.3 <.001 MK 0.90 0.83–0.97 0.567 ↓ 94.1 78.3 <.001 KFA 0.55 0.42–0.68 0.807 ↑ 63.2 84.2 .73 MKT 0.91 0.84–0.98 0.619 ↓ 88.2 84.8 <.001 SMT LMDc 0.62 0.48–0.75 2.967 ↑ 73.9 58.8 .009 TMDc 0.87 0.79–0.95 0.528 ↑ 78.3 91.2 <.001 µFA 0.91 0.84–0.98 0.535 ↓ 91.2 84.8 <.001 µFA3 0.91 0.84–0.98 0.153 ↓ 91.2 84.8 <.001 MMDc 0.85 0.77–0.94 1.369 ↑ 67.4 91.2 <.001 INVF 0.91 0.84–0.98 0.339 ↓ 91.2 84.8 <.001 IDc 0.72 0.61–0.83 2.006 ↑ 69.6 70.6 <.001 ETMDc 0.77 0.66–0.87 1.301 ↑ 76.1 67.6 <.001 EMMDc 0.65 0.53–0.78 1.502 ↑ 67.4 67.6 .01 Note:—T2LFM indicates T2-FLAIR mismatch; FA, fractional anisotropy; MD, mean diffusivity; KFA, kurtosis fractional anisotropy; LMD, longitudinal microscopic diffusivity; µFA3, microscopic fractional anisotropy to the third power; MMD, microscopic mean diffusivity; ID, intrinsic diffusivity, ETMD, extra-neurite transverse microscopic diffusivity; EMMD, extra-neurite microscopic mean diffusivity.
↵a Optimal cutoff levels to predict IDH type were assessed by the Youden index. Cutoffs were evaluated by sensitivity and specificity. An upward arrow (↑) indicates a positive correlation, in which values above the cutoff point predict an IDH-mutant glioma, whereas a downward arrow (↓) indicates a negative correlation, in which values below the cutoff point predict an IDH-mutant glioma. P values were computed by comparing the AUC against chance performance.
↵b Binary variable, indicating the presence or absence of the feature.
↵c Units in mm2/s × 10−3.
- Table 3:
Comparative diagnostic accuracy of investigated parameters differentiating glioma gradesa
Models and Parameters AUC 95% CI Optimal Cutoff Sensitivity (%) Specificity (%) P Value Nondiffusion Enhancementb 0.76 0.65–0.87 0.5 ↑ 81.4 70.2 <.001 Necrosisb 0.81 0.71–0.91 0.5 ↑ 67.4 94.6 <.001 T2LFMb 0.58 0.46–0.71 0.5 ↓ 16.2 100 .006 Hemorrhageb 0.73 0.61–0.84 0.5 ↑ 51.2 94.6 <.001 ADC ADCc 0.88 0.80–0.96 1.217 ↓ 75.7 88.4 <.001 DTI FA 0.64 0.52–0.77 0.174 ↑ 46.5 86.5 .01 MDc 0.88 0.80–0.95 1.134 ↓ 83.8 76.7 <.001 DKI AK 0.93 0.87–1.00 0.494 ↑ 97.7 75.7 <.001 RK 0.93 0.88–1.00 0.627 ↑ 83.7 91.9 <.001 MK 0.93 0.88–1.00 0.539 ↑ 93.0 81.1 <.001 KFA 0.53 0.40–0.67 0.219 ↓ 78.4 44.2 .85 MKT 0.94 0.88–0.99 0.564 ↑ 88.4 86.5 <.001 SMT LMDc 0.66 0.53–0.79 2.967 ↓ 83.8 55.8 <.001 TMDc 0.91 0.84–0.97 2.958 ↓ 91.9 79.1 <.001 µFA 0.94 0.88–0.99 0.509 ↑ 90.7 86.5 <.001 µFA3 0.94 0.88–0.99 0.132 ↑ 90.7 86.5 <.001 MMDc 0.90 0.83–0.97 1.311 ↓ 89.2 76.7 <.001 INVF 0.94 0.88–0.99 0.319 ↑ 90.7 86.5 <.001 IDc 0.77 0.67–0.88 1.946 ↓ 83.8 65.1 <.001 ETMDc 0.82 0.72–0.91 1.208 ↓ 94.6 60.5 <.001 EMMDc 0.72 0.60–0.84 1.502 ↓ 75.7 67.4 <.001 Note:—T2LFM indicates T2 FLAIR mismatch; FA, fractional anisotropy; MD, mean diffusivity; KFA, kurtosis fractional anisotropy; LMD, longitudinal microscopic diffusivity; µFA3, microscopic fractional anisotropy to the third power; MMD, microscopic mean diffusivity; ID, intrinsic diffusivity; ETMD, extraneurite transverse microscopic diffusivity; EMMD, extraneurite microscopic mean diffusivity.
↵a Optimal cutoff levels to predict glioma grade (low-grade versus high-grade) were assessed by the Youden index. Cutoffs were evaluated by sensitivity and specificity. An upward arrow (↑) indicates a positive correlation, in which values above the cutoff point predict a high-grade glioma, whereas a downward arrow (↓) indicates a negative correlation, in which values below the cutoff point predict a high-grade glioma. P values were computed by comparing the AUC against chance performance.
↵b Binary variable, indicating the presence or absence of the feature.
↵c Units mm2/s × 10−3.
Jump to section
Related Articles
Cited By...
- No citing articles found.