MST-based local boundary leakage and global topology divergence, fused by task complexity, rank SSL 3D medical encoders for segmentation without fine-tuning, beating prior TE metrics by 0.36 weighted Kendall τ at 56× speed.
Training dataset for hntsmrg 2024 challenge.https://zenodo.org/doi/10.5281/zenodo.11199559, 2024
1 Pith paper cite this work, alongside 8 external citations. Polarity classification is still indexing.
1
Pith paper citing it
8
external citations · OpenAlex
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Topology-Driven Transferability Estimation for 3D Medical Vision Foundation Models
MST-based local boundary leakage and global topology divergence, fused by task complexity, rank SSL 3D medical encoders for segmentation without fine-tuning, beating prior TE metrics by 0.36 weighted Kendall τ at 56× speed.