SegMamba & GliomaSAM3-MoE: Complete Reproduction Package

This repository contains everything needed to reproduce our brain tumor segmentation experiments on BraTS 2023 dataset, including source code, pre-trained weights, and evaluation results.

Repository Structure

.
├── source_code/
│   ├── gliomasam3_moe/        # GliomaSAM3-MoE source code
│   ├── SegMamba/              # SegMamba source code (with mamba/monai)
│   └── sam3/                  # SAM3 dependency module
│
├── pretrained_weights/
│   └── sam3.pt                # SAM3 pretrained weights (3.3GB)
│
├── gliomasam3_moe/
│   ├── checkpoints/           # GliomaSAM3-MoE trained weights
│   │   ├── ckpt_step2000.pt
│   │   ├── ckpt_step2600.pt
│   │   └── ckpt_step3000.pt   # Best checkpoint
│   ├── configs/
│   │   └── train.yaml         # Training configuration
│   ├── eval_results/
│   │   ├── table4_et_absent.json   # ET presence classification results
│   │   └── table7_boundary_dice.json # Boundary-band Dice results
│   └── vis_res/               # Visualization results
│       ├── method_comparison/ # Side-by-side comparisons
│       ├── boundary/          # Boundary analysis figures
│       ├── moe_routing/       # MoE routing visualizations
│       └── ...
│
├── segmamba/
│   ├── checkpoints/           # SegMamba trained weights
│   │   ├── tmp_model_ep599_0.8295.pt
│   │   └── tmp_model_ep799_0.8498.pt  # Best checkpoint (Dice=0.8498)
│   └── prediction_results/
│       ├── segmamba_brats23_ep799/    # Prediction NIfTI files
│       └── result_metrics/            # Evaluation metrics
│
└── README.md

Model Performance

GliomaSAM3-MoE (ckpt_step3000)

Boundary-band Dice (3-voxel band):

Region	Dice
WT	0.789 ± 0.057
TC	0.766 ± 0.154
ET	0.697 ± 0.161
Mean	0.750

ET Presence Classification:

Metric	Value
AUROC	0.896
Accuracy	0.795
Sensitivity	0.792
Specificity	1.000

SegMamba (ep799)

• Mean Dice: 0.8498
• Trained for 800 epochs on BraTS 2023

Quick Start: Reproduction

1. Download this repository

# Clone the dataset
git clone https://huggingface.co/datasets/ChipYTY/segmamba
cd segmamba

2. Prepare BraTS 2023 Data

Download BraTS 2023 GLI Challenge data from Synapse and preprocess:

cd source_code/SegMamba
python 2_preprocessing_mri.py --input_dir /path/to/BraTS2023 --output_dir /path/to/processed

3. Run GliomaSAM3-MoE Inference

cd source_code/gliomasam3_moe

# Set SAM3 path
export PYTHONPATH=/path/to/source_code/sam3:$PYTHONPATH
export SAM3_CKPT=/path/to/pretrained_weights/sam3.pt

# Run inference
python infer.py \
    --config configs/train.yaml \
    --checkpoint /path/to/gliomasam3_moe/checkpoints/ckpt_step3000.pt \
    --data_dir /path/to/processed \
    --output_dir ./predictions

4. Run SegMamba Inference

cd source_code/SegMamba

python 4_predict.py \
    --checkpoint /path/to/segmamba/checkpoints/tmp_model_ep799_0.8498.pt \
    --data_dir /path/to/processed \
    --output_dir ./predictions

Usage

Loading GliomaSAM3-MoE

import torch

# Load checkpoint
ckpt = torch.load("gliomasam3_moe/checkpoints/ckpt_step3000.pt", map_location="cpu")

# Model state dict is in ckpt["model"]
model.load_state_dict(ckpt["model"])

Loading SegMamba

import torch

# Load checkpoint
ckpt = torch.load("segmamba/checkpoints/tmp_model_ep799_0.8498.pt", map_location="cpu")

# Model state dict
model.load_state_dict(ckpt["model"])

Data

Models were trained and evaluated on BraTS 2023 GLI Challenge dataset.

• Download: Synapse BraTS 2023
• Preprocessing: Use source_code/SegMamba/2_preprocessing_mri.py

Requirements

• Python 3.10+
• PyTorch 2.0+
• CUDA 11.8+ (for SegMamba's Mamba CUDA kernels)
• See source_code/gliomasam3_moe/requirements.txt for full list

Citation

If you use these models, please cite the relevant papers.

License

Please refer to the original model repositories for licensing information.