Research map / Video / Video segmentation and tracking
Δt-Mamba3D: A Time‑Aware Spatio‑Temporal State‑Space Model for Breast Cancer Risk Prediction
Zhengbo Zhou, Dooman Arefan, Margarita Louise Zuley, Shandong Wu
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 true inter-visit Δt modulation to Mamba (Δt-Mamba vs baseline Mamba) (ablation) | Mamba · CSAW (max 4 prior exams) | C-index: 0.701 -> 0.716 | Improved |
| add multi-scale 3D neighborhood fusion on top of Δt-Mamba (ablation) | Δt-Mamba · CSAW (max 4 prior exams) | C-index: 0.716 -> 0.742 | Improved |
| add multi-scale 3D neighborhood fusion on top of Δt-Mamba (ablation) | Δt-Mamba · Independent dataset (max 8 prior exams) | C-index: 0.717 -> 0.738 | Improved |
| add spatial-only (1,3,3) 3D kernel to baseline Mamba (ablation) | Mamba · CSAW (max 4 prior exams) | C-index: 0.701 -> 0.705 | Improved |
| add spatio-temporal (3,3,3) 3D kernel alone to baseline Mamba (ablation) | Mamba · CSAW (max 4 prior exams) | C-index: 0.701 -> 0.702 | No clear effect |
| combine spatial and spatio-temporal kernels (ablation) | Mamba3D · CSAW (max 4 prior exams) | C-index: 0.705 -> 0.713 | Improved |
| use inter-slice scan with Δt at every hop (ablation) | Δt-Mamba · CSAW (max 4 prior exams) | C-index: 0.716 -> 0.682 | Hurt |
| use Δt-Mamba3D instead of ContiFormer | longitudinal risk prediction · CSAW (max 4 prior exams) | C-index: 0.659 -> 0.742 | Improved |
| use Δt-Mamba3D instead of Transformer for efficiency | single-layer model · Single-layer efficiency (512 tokens) | Peak tokens/s (x10^6): 11.9 -> 59.3 | Improved |
| use Δt-Mamba3D instead of GRU-Δt for speed | single-layer model · Single-layer efficiency (512 tokens) | Peak tokens/s (x10^6): 79.0 -> 59.3 | Hurt |
Closest papers by meaning
- Multimodal Disease Progression Modeling via Spatiotemporal Disentanglement and Multiscale Alignment (NeurIPS 2025 · similarity 0.86)
- Time-to-Event Pretraining for 3D Medical Imaging (ICLR 2025 · similarity 0.85)
- Autoregressive Sequence Modeling for 3D Medical Image Representation (AAAI 2025 · similarity 0.85)
- Spatial-Mamba: Effective Visual State Space Models via Structure-Aware State Fusion (ICLR 2025 · similarity 0.84)
- Integrating Sequence and Image Modeling in Irregular Medical Time Series Through Self-Supervised Learning (AAAI 2025 · similarity 0.83)
- Unlocking the Power of Spatial and Temporal Information in Medical Multimodal Pre-training (ICML 2024 · 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.
