Copyright: ©Author(s) 2026.
World J Gastrointest Oncol. Jul 15, 2026; 18(7): 120437
Published online Jul 15, 2026. doi: 10.4251/wjgo.120437
Published online Jul 15, 2026. doi: 10.4251/wjgo.120437
Table 3 Ablation experimental results on the independent test set
| Model configuration | T-stage AUC | N-stage AUC | T-staging accuracy | N-staging accuracy | Weighted Kappa coefficient |
| Single-path backbone network | 0.852 | 0.861 | 0.835 | 0.871 | 0.784 |
| Removal of dual-path fusion | 0.873 | 0.879 | 0.859 | 0.891 | 0.812 |
| Removal of cross-attention mechanism | 0.881 | 0.886 | 0.864 | 0.902 | 0.828 |
| Removal of feature pyramid network | 0.892 | 0.898 | 0.876 | 0.918 | 0.846 |
| Complete model (the method proposed herein) | 0.907 | 0.912 | 0.898 | 0.949 | 0.883 |
- Citation: Zhao J, Du LJ, Liu Y, Zhu DD, Wang HQ, Shen MK, Wang LY, Wang HY. Development and clinical application of an ultrasound-based deep learning model for preoperative staging of colorectal cancer. World J Gastrointest Oncol 2026; 18(7): 120437
- URL: https://www.wjgnet.com/1948-5204/full/v18/i7/120437.htm
- DOI: https://dx.doi.org/10.4251/wjgo.120437