MedPointS-SEG: Medical Point Cloud Segmentation Dataset

This repository contains the MedPointS-SEG dataset, a medical point cloud segmentation dataset presented in the paper Hierarchical Feature Learning for Medical Point Clouds via State Space Model.

• Project Page: https://flemme-docs.readthedocs.io/en/latest/medpoints.html
• Code: https://github.com/wlsdzyzl/flemme

Paper Abstract

Deep learning-based point cloud modeling has been widely investigated as an indispensable component of general shape analysis. Recently, transformer and state space model (SSM) have shown promising capacities in point cloud learning. However, limited research has been conducted on medical point clouds, which have great potential in disease diagnosis and treatment. This paper presents an SSM-based hierarchical feature learning framework for medical point cloud understanding. Specifically, we down-sample input into multiple levels through the farthest point sampling. At each level, we perform a series of k-nearest neighbor (KNN) queries to aggregate multi-scale structural information. To assist SSM in processing point clouds, we introduce coordinate-order and inside-out scanning strategies for efficient serialization of irregular points. Point features are calculated progressively from short neighbor sequences and long point sequences through vanilla and group Point SSM blocks, to capture both local patterns and long-range dependencies. To evaluate the proposed method, we build a large-scale medical point cloud dataset named MedPointS for anatomy classification, completion, and segmentation. Extensive experiments conducted on MedPointS demonstrate that our method achieves superior performance across all tasks. The dataset is available at this https URL . Code is merged to a public medical imaging platform: this https URL .

Dataset Description

This is the medical point cloud segmentation dataset from MedPointS, where data is the input point cloud, and label is the class label.

Each point cloud has been normalized and sub-sampled to 4096 points.

Sample Usage

The MedPointS-SEG dataset can be used with the Flemme platform for training and evaluating point cloud segmentation models.

First, ensure you have the Flemme platform set up by following the instructions in its GitHub repository. After setting up Flemme, you can train and test a segmentation model using the provided configuration files, for example:

## segmentation
train_flemme --config /path/to/project/flemme/resources/pcd/medpoints/seg/train_pointmamba2knn_sem.yaml
test_flemme --config /path/to/project/flemme/resources/pcd/medpoints/seg/test_pointmamba2knn_sem.yaml

Citation

If you find our project helpful, please consider to cite the following work:

@misc{zhang2025hierarchicalfeaturelearningmedical,
      title={Hierarchical Feature Learning for Medical Point Clouds via State Space Model},
      author={Guoqing Zhang and Jingyun Yang and Yang Li},
      year={2025},
      eprint={2504.13015},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2504.13015},
}

---

dataset_info: features: - name: partial sequence: sequence: float32 - name: label sequence: sequence: float32 splits: - name: train num_bytes: 856481800 num_examples: 1025 download_size: 60709044 dataset_size: 856481800 configs: - config_name: default data_files: - split: train path: data/train-*