Copyright
©The Author(s) 2025.
World J Radiol. Dec 28, 2025; 17(12): 112911
Published online Dec 28, 2025. doi: 10.4329/wjr.v17.i12.112911
Published online Dec 28, 2025. doi: 10.4329/wjr.v17.i12.112911
Table 1 Magnetic resonance imaging scan sequences and parameters
| Sequence | Layer thickness (mm) | Field (mm2) | TR/TE (ms) | Flip angle (°) | Matrix size (mm2) |
| RTr Ax fs T2 | 6 | 320 × 320 - 380 × 380 | 2609/97 | 110 | 384 × 384 |
| BH Ax LAVA +C | 5 | 320 × 320 - 360 × 360 | 4/2 | 12 | 224 × 192 |
Table 2 Comparison of the clinical characteristics between the training and validation groups, n (%)
| Clinical features | Training group (n = 87) | Validation group (n = 38) | P value | ||
| Nonpoorly differentiated HCC (n = 73) | Poorly differentiated HCC (n = 14) | Nonpoorly differentiated HCC (n = 32) | Poorly differentiated HCC (n = 6) | ||
| Age (years) | 55.00 (48.00, 65.00) | 52.50 (44.50, 64.00) | 53.50 (47.75, 63.25) | 58.00 (37.00, 64.75) | 0.621 |
| NLR | 2.56 (2.07, 3.33) | 3.09 (2.50, 3.75) | 2.86 (1.85, 5.66) | 4.67 (2.88, 6.95) | 0.185 |
| PLR | 94.83 (70.66, 134.21) | 129.27 (120.54, 177.77) | 128.36 (75.52, 194.57) | 155.89 (140.21, 222.43) | 0.063 |
| LMR | 3.98 (2.97, 4.58) | 3.03 (2.43, 3.58) | 3.53 (1.98, 4.44) | 1.94 (1.41, 2.68) | 0.041 |
| Maximum tumor diameter (cm) | 4.60 (3.10, 6.10) | 8.30 (6.30, 10.57) | 5.65 (4.52, 7.73) | 9.20 (6.10, 10.43) | 0.052 |
| AST (U/L) | 40.00 (30.00, 68.00) | 42.50 (29.50, 66.50) | 44.00 (38.00, 64.25) | 81.50 (65.25, 86.50) | 0.102 |
| ALT (U/L) | 32.00 (21.00, 59.00) | 45.00 (24.75, 58.25) | 34.50 (23.75, 49.50) | 57.00 (26.00, 74.50) | 0.906 |
| GGT (U/L) | 62.00 (30.00, 112.00) | 81.50 (25.75, 254.75) | 87.00 (43.80, 150.75) | 363.00 (131.00, 600.25) | 0.058 |
| NEU (109/L) | 3.49 (2.78, 4.33) | 5.29 (3.34, 7.73) | 4.02 (2.43, 6.06) | 4.87 (4.38, 5.61) | 0.399 |
| LYMPH (109/L) | 1.34 (1.05, 1.72) | 1.34 (1.10, 2.08) | 1.27 (0.88, 1.49) | 1.25 (0.91, 1.50) | 0.125 |
| PLT (109/L) | 136.00 (100.00, 180.00) | 193.50 (137.75, 224.50) | 156.50 (109.25, 203.00) | 198.00 (190.50, 203.25) | 0.251 |
| MONO (109/L) | 0.39 (0.30, 0.46) | 0.42 (0.38, 0.72) | 0.40 (0.31, 0.48) | 0.69 (0.59, 0.80) | 0.263 |
| A/G ratio | 1.29 (1.13, 1.53) | 1.18 (0.96, 1.41) | 1.31 (1.21, 1.44) | 1.08 (1.01, 1.10) | 0.646 |
| Sex | 0.214 | ||||
| Female | 6 (8.22) | 3 (21.43) | 6 (18.75) | 1 (16.67) | |
| Male | 67 (91.78) | 11 (78.57) | 26 (81.25) | 5 (83.33) | |
| Portal vein tumor thrombus | 0.400 | ||||
| No | 63 (86.30) | 9 (64.29) | 27 (84.38) | 2 (33.33) | |
| Yes | 10 (13.70) | 5 (35.71) | 5 (15.62) | 4 (66.67) | |
| Lymph node metastasis | 0.674 | ||||
| No | 63 (86.30) | 10 (71.43) | 31 (96.88) | 2 (33.33) | |
| Yes | 10 (13.70) | 4 (28.57) | 1 (3.12) | 4 (66.67) | |
| Liver cirrhosis | 0.385 | ||||
| No | 14 (19.18) | 7 (50.00) | 11 (34.38) | 1 (16.67) | |
| Yes | 59 (80.82) | 7 (50.00) | 21 (65.62) | 5 (83.33) | |
| HBV | 0.336 | ||||
| No | 8 (10.96) | 7 (50.00) | 2 (6.25) | 2 (33.33) | |
| Yes | 65 (89.04) | 7 (50.00) | 30 (93.75) | 4 (66.67) | |
| AFP (ng/mL) | 0.276 | ||||
| < 400 | 48 (65.75) | 9 (64.29) | 18 (56.25) | 3 (50.00) | |
| ≥ 400 | 25 (34.25) | 5 (35.71) | 14 (43.75) | 3 (50.00) | |
| ALBI grade | 0.642 | ||||
| 1 | 38 (52.05) | 5 (35.71) | 14 (43.75) | 4 (66.67) | |
| 2 | 15 (20.55) | 6 (42.86) | 10 (31.25) | 2 (33.33) | |
| 3 | 20 (27.40) | 3 (21.43) | 8 (25.00) | 0 (0.00) | |
Table 3 Screened optimal radiomic features
| MRI sequence | Feature type | Feature name |
| FS-T2WI | wavelet-HLL_first order | Kurtosis |
| wavelet-HLL_glcm | Idn | |
| wavelet-HLL_gldm | Large dependence high gray level emphasis | |
| AP | Original_shape | Sphericity |
| Original_glcm | Cluster prominence | |
| Original_glszm | Large area high gray level emphasis | |
| wavelet-LLL_glcm | Cluster prominence | |
| PVP | Original_shape | Sphericity |
| log-sigma-2-5-mm-3D_ gldm | Small dependence high gray level emphasis | |
| wavelet-HHL_glszm | Large area emphasis | |
| wavelet-HHL_glszm | Small area emphasis | |
| wavelet-LLL_glcm | Cluster prominence |
Table 4 Results of univariate and multivariate logistic regression analyses
| Clinical factors | Univariate logistic regression analysis | Multivariate logistic regression analysis | ||
| OR (95%CI) | P value | OR (95%CI) | P value | |
| Age (years) | 0.980 (0.934-1.028) | 0.406 | - | - |
| Sex | 0.328 (0.071-1.510) | 0.152 | - | - |
| NLR | 1.009 (0.935-1.088) | 0.826 | - | - |
| PLR | 1.005 (0.997-1.012) | 0.209 | - | - |
| LMR | 0.675 (0.439-1.039) | 0.074 | - | - |
| Maximum tumor diameter (cm) | 1.232 (1.051-1.444) | 0.0101 | 1.364 (1.065-1.746) | 0.0141 |
| Portal vein tumor thrombus | 3.500 (0.972-12.597) | 0.055 | - | - |
| Lymph node metastasis | 2.520 (0.661-9.603) | 0.176 | - | - |
| HBV | 0.106 (0.029-0.391) | < 0.0011 | 0.120 (0.020-0.729) | 0.0211 |
| Liver cirrhosis | 0.237 (0.072-0.787) | 0.0191 | 0.326 (0.053-2.020) | 0.228 |
| AFP (ng/mL) | 1.067 (0.323-3.525) | 0.916 | - | - |
| PLT (109/L) | 1.007 (1.001-1.014) | 0.0351 | 0.994 (0.984-1.005) | 0.296 |
| LYMPH (109/L) | 1.364 (0.614-3.030) | 0.446 | - | - |
| NEU (109/L) | 1.140 (0.981-1.325) | 0.088 | - | - |
| MONO (109/L) | 2.312 (1.324-4.035) | 0.0031 | 2.488 (1.175-5.271) | 0.0171 |
| AST (U/L) | 1.004 (0.998-1.010) | 0.181 | - | - |
| ALT (U/L) | 1.004 (0.999-1.008) | 0.150 | - | - |
| GGT (U/L) | 1.002 (0.999-1.005) | 0.211 | - | - |
| A/G ratio | 0.198 (0.022-1.779) | 0.148 | - | - |
| ALBI grade | - | 0.219 | - | - |
| ALBI grade 1 | 0.877 (0.19-4.052) | 0.867 | - | - |
| ALBI grade 2 | 2.667 (0.572-12.428) | 0.212 | - | - |
Table 5 Predictive performance of each model
| Group | Model | AUC value (95%CI) | Accuracy | Sensitivity | Specificity | Positive predictive value | Negative predictive value |
| Training group | FS-T2WI | 0.761 (0.562-0.857) | 0.862 | 0.214 | 0.986 | 0.750 | 0.867 |
| AP | 0.870 (0.714-0.918) | 0.897 | 0.500 | 0.973 | 0.778 | 0.910 | |
| PVP | 0.868 (0.714-0.959) | 0.908 | 0.500 | 0.986 | 0.875 | 0.911 | |
| Radiomics | 0.917 (0.857-0.959) | 0.943 | 0.714 | 0.986 | 0.909 | 0.947 | |
| Clinical | 0.869 (0.643-0.973) | 0.885 | 0.429 | 0.973 | 0.750 | 0.899 | |
| Combined | 0.941 (0.857-0.945) | 0.942 | 0.714 | 0.986 | 0.909 | 0.947 | |
| Validation group | FS-T2WI | 0.724 (0.625-0.833) | 0.816 | 0.333 | 0.906 | 0.400 | 0.879 |
| AP | 0.802 (0.686-1.000) | 0.763 | 0.333 | 0.844 | 0.286 | 0.871 | |
| PVP | 0.797 (0.688-1.000) | 0.763 | 0.167 | 0.875 | 0.200 | 0.848 | |
| Radiomics | 0.901 (0.833-0.906) | 0.895 | 0.667 | 0.938 | 0.667 | 0.938 | |
| Clinical | 0.865 (0.594-1.000) | 0.895 | 0.500 | 0.969 | 0.750 | 0.912 | |
| Combined | 0.932 (0.812-1.000) | 0.895 | 0.667 | 0.938 | 0.667 | 0.938 |
- Citation: Shi Y, Zhang P, Li L, Yang HM, Li ZM, Zheng J, Yang L. Interpretable model based on multisequence magnetic resonance imaging radiomics for predicting the pathological grades of hepatocellular carcinomas. World J Radiol 2025; 17(12): 112911
- URL: https://www.wjgnet.com/1949-8470/full/v17/i12/112911.htm
- DOI: https://dx.doi.org/10.4329/wjr.v17.i12.112911
