MELA Dataset: A Benchmark for Mediastinal Lesion Analysis (Validation Set and Annotation)

Shuang Song, Rui Xu, Yong Luo, Bo Du, Jiancheng Yang, Kaiming Kuang, Yunlang She, Mengmeng ZhaoZenodo 2022 · DOI · OpenAlex

Abstract

MELA dataset is a benchmark for developing algorithms on mediastinal lesion analysis. We hope this large-scale dataset could facilitate the research and application of automatic mediastinal lesion detection and diagnosis. MELA dataset contains 1100 CT scans collected from patients with one or more lesions in the mediastinum. The MELA dataset is split into a subset of 770 CT scans for training, a subset of 110 CT scans for validation, and a test set of 220 CT scans for evaluation. This is the Validation Set and Annotation of MELA dataset, including 110 CTs and the annotations of the whole training set and validation set. Files include: Val.zip: 110 CTs in NII format (nii.gz). mela_train_val_annotations.csv: bounding box annotations in voxel coordinates for mediastinal lesions. `public_id: anonymous patient ID to match images and annotations.<br> `coordX, coordY, coordZ: coordinates of the center of annotated bounding box.<br> `x_length, y_length, z_length: the length of the bounding box in three dimensions.
BibTeX
@inproceedings{song2022mela,
  author = {Shuang Song and Rui Xu and Yong Luo and Bo Du and Jiancheng Yang and Kaiming Kuang and Yunlang She and Mengmeng Zhao},
  title = {{MELA Dataset: A Benchmark for Mediastinal Lesion Analysis (Validation Set and Annotation)}},
  booktitle = {Zenodo},
  year = {2022},
  doi = {10.5281/zenodo.6597131},
}

Loading related content...

Figure