Research map / Detection and segmentation / CNN features
SPHERE: From MRI Sampling Mechanisms to Spatial Priors for Generalizable Brain Tumor Segmentation
JiaCheng Lu, Shiyu Zhang, Hui Ding, junhui xin, Guoping Huo
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add DVF prior with multiply fusion (ablation) | Base Swin-RetNet · BraTS 2019 (train BraTS 2023) | Dice avg: 89.79 -> 91.03 | Improved |
| add recovered theta prior (ablation) | Base Swin-RetNet · BraTS 2019 (train BraTS 2023) | Dice avg: 89.79 -> 90.44 | Improved |
| add auxiliary ViT branch (ablation) | Base Swin-RetNet · BraTS 2019 (train BraTS 2023) | Dice avg: 89.79 -> 89.63 | No clear effect |
| use DVF instead of artifact magnitude prior (multiply fusion) (ablation) | Artifact prior · BraTS 2019 (train BraTS 2023) | Dice avg: 89.93 -> 91.03 | Improved |
| use DVF instead of gradient energy prior (multiply fusion) (ablation) | Gradient prior · BraTS 2019 (train BraTS 2023) | Dice avg: 90.45 -> 91.03 | Improved |
| use explicit theta->DVF construction instead of end-to-end DVF (ablation) | E2E-DVF · BraTS 2019 (train BraTS 2023) | Dice avg: 90.85 -> 91.27 | Improved |
| keep DVF spatial structure instead of shuffling it (ablation) | Shuffle-DVF · BraTS 2019 (train BraTS 2023) | Dice avg: 89.36 -> 91.27 | Improved |
| use RetNet fusion instead of multiply fusion (ablation) | DVF prior · BraTS 2019 (train BraTS 2023) | Dice avg: 91.03 -> 91.27 | Improved |
| use RetNet fusion instead of SE fusion (ablation) | DVF prior · BraTS 2019 (train BraTS 2023) | Dice avg: 90.72 -> 91.27 | Improved |
| use RetNet fusion instead of conv fusion (ablation) | DVF prior · BraTS 2019 (train BraTS 2023) | Dice avg: 90.94 -> 91.27 | Improved |
| add DVF prior with RetNet fusion (ablation) | Base Swin-RetNet · BraTS 2023 | Dice avg: 92.25 -> 92.87 | Improved |
| combine spatial ViT branch and frequency branch (ablation) | SPHERE-Recover · DICOM test set | MAE of Nf: 0.09428 -> 0.08531 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- M-Net — M-Net: MRI Brain Tumor Sequential Segmentation Network via Mesh-Cast (ICCV 2025) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- M-Net: MRI Brain Tumor Sequential Segmentation Network via Mesh-Cast (ICCV 2025 · similarity 0.85)
- Incomplete Multi-modal Brain Tumor Segmentation via Learnable Sorting State Space Model (CVPR 2025 · similarity 0.85)
- Adaptive Sampling of k-Space in Magnetic Resonance for Rapid Pathology Prediction (ICML 2024 · similarity 0.85)
- Physics-Regularized Multi-Modal Image Assimilation for Brain Tumor Localization (NeurIPS 2024 · similarity 0.84)
- Sequential Information Bottleneck Fusion: Towards Robust and Generalizable Multi-Modal Brain Tumor Segmentation (ICLR 2026 · similarity 0.84)
- KMD: Koopman Multi-modality Decomposition for Generalized Brain Tumor Segmentation under Incomplete Modalities (CVPR 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
