--- license: cc-by-nd-4.0 size_categories: - n>1T tags: - medical dataset_info: features: - name: ICT dtype: image - name: LDCT_Low dtype: image - name: LDCT_Mid dtype: image - name: LDCT_High dtype: image - name: LACT_Low dtype: image - name: LACT_Mid dtype: image - name: LACT_High dtype: image - name: SVCT_Low dtype: image - name: SVCT_Mid dtype: image - name: SVCT_High dtype: image splits: - name: train_previews num_bytes: 62199112.0 num_examples: 44 - name: test_previews num_bytes: 16108271.0 num_examples: 11 download_size: 153191938 dataset_size: 78307383.0 configs: - config_name: default data_files: - split: train_previews path: data/train_previews-* - split: test_previews path: data/test_previews-* --- # SimNICT: Simulated Non-Ideal measurement CT Dataset **SimNICT** is the first comprehensive dataset for training universal non-ideal measurement CT (NICT) enhancement models, containing simulated low-dose, limited-angle, and sparse-view CT from different body regions. We release the **SimNICT Dataset** (823 GB, 8 datasets) for comprehensive NICT research, and provide **SimNICT-AMOS-Sample** (78 MB) for quick exploration and prototyping. **šŸ’” Recommendation**: Start with [SimNICT-AMOS-Sample Dataset](#part-2-simnict-amos-sample) for initial exploration and prototyping, then download specific datasets from the [SimNICT Dataset](#part-1-simnict-dataset) based on your research needs. --- ## Part 1: SimNICT Dataset | Dataset | Volumes | Body Regions | License | Download Link | | ------------------------------- | ------- | ------------ | --------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------- | | **AMOS** | 500 | Abdomen | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | [simnict-amos](https://archive.org/details/simnict-amos) | | **COVID-19-NY-SBU** | 459 | Chest | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | [simnict-covid-19-ny-sbu](https://archive.org/details/simnict-covid-19-ny-sbu) | | **CT Images in COVID-19** | 771 | Chest | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | [simnict-ct-images-in-covid-19](https://archive.org/details/simnict-ct-images-in-covid-19) | | **CT_COLONOGRAPHY** | 1,730 | Abdomen | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | [simnict-ct-colonography](https://archive.org/details/simnict-ct-colonography) | | **LNDb** | 294 | Chest | [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/) | [simnict-lndb](https://archive.org/details/simnict-lndb) | | **LUNA** | 888 | Chest | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | [simnict-luna](https://archive.org/details/simnict-luna) | | **MELA** | 1,100 | Chest | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | [simnict-mela](https://archive.org/details/simnict-mela) | | **STOIC** | 2,000 | Chest | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | [simnict-stoic](https://archive.org/details/simnict-stoic) | | **AutoPET** | 1,014 | Whole-body | [NIH Controlled Data Access Policy](https://www.cancerimagingarchive.net/nih-controlled-data-access-policy/) | - | | **HECKTOR22** | 882 | Head, neck | [Custom Research License](https://hecktor.grand-challenge.org/Participation/) | - | *Note: AutoPET and HECKTOR22 datasets are not publicly available due to licensing restrictions.* ### Dataset Overview The **SimNICT dataset** is a large-scale medical imaging dataset containing: - **šŸ“Š 9,513 CT volumes** from 10 medical imaging datasets (2 out of 10 datasets are not open-source due to licensing restrictions) - **šŸ”¬ 3 NICT types**: Low-dose CT (LDCT), Sparse-view CT (SVCT), Limited-angle CT (LACT) - **āš™ļø Randomized parameters**: - **SVCT**: Views randomly sampled from 15-360 range - **LACT**: Angular range randomly sampled from 75Ā°-270Ā° - **LDCT**: Dose levels randomly sampled from 5%-75% range - **šŸ’¾ Total size**: ~823 GB - **šŸ“ File format**: NIfTI (.nii.gz), 16-bit, gzip compressed ### Data Release Strategy The SimNICT dataset provides **preprocessed ICT data** with [**NICT simulation code**](https://huggingface.co/datasets/YutingHe-list/SimNICT/resolve/main/simnict_generator.py): **1. šŸ” Preprocessed ICT Data** The preprocessed ICT data of SimNICT dataset is hosted on [Internet Archive](https://archive.org/search.php?query=simnict), ensuring stable access for the global research community. You can download data through the [dataset table above](#part-1-simnict-dataset) or execute batch download scripts below: ```bash # Download batch download script wget https://huggingface.co/datasets/YutingHe-list/SimNICT/blob/main/simnict_download.py # Download all datasets (~823 GB) python simnict_download.py --all --output_dir ./data # Download specific datasets python simnict_download.py --datasets AMOS LUNA --output_dir ./data ``` **2. āš™ļø NICT Simulation Code** After downloading preprocessed ICT data, generate NICT data with the following simulation code to construct complete SimNICT dataset: ```python # Download simulation code wget https://huggingface.co/datasets/YutingHe-list/SimNICT/blob/main/simnict_generator.py # Configure paths and run python simnict_generator.py ``` The simulation code uses advanced physics-based modeling with **ODL** (Operator Discretization Library) and **ASTRA Toolbox** for accurate CT reconstruction simulation. --- ## Part 2: SimNICT-AMOS-Sample **SimNICT-AMOS-Sample** is a preview subset of SimNICT dataset for quick exploration and prototyping. ### Sample Dataset Specifications - **šŸ“‚ Source**: Selected from AMOS dataset (part of SimNICT) - **šŸ“Š Content**: 55 CT volumes (44 train + 11 test) - **šŸ”¬ Coverage**: 3 NICT types Ɨ 3 fixed severity levels (different from SimNICT dataset's randomized parameters) - **šŸ’¾ Size**: ~78 MB (1000Ɨ smaller than SimNICT dataset) - **šŸš€ Format**: Preprocessed and optimized for Hugging Face platform ### Quick Start with Sample Dataset ```python from datasets import load_dataset # Load the preview sample dataset dataset = load_dataset("YutingHe-list/SimNICT") sample = dataset["train_previews"][0] # Access different NICT simulations ict_image = sample["ICT"] # Ground truth ldct_low = sample["LDCT_Low"] # Low-dose simulation svct_mid = sample["SVCT_Mid"] # Sparse-view simulation lact_high = sample["LACT_High"] # Limited-angle simulation ``` ### NICT Simulation Parameters (Sample Dataset Only) | Type | Low | Mid | High | | -------------- | ----------- | ----------- | ----------- | | **LDCT** | Iā‚€=1Ɨ10āµ | Iā‚€=1Ɨ10ā“ | Iā‚€=1Ɨ10Ā³ | | **SVCT** | 120 views | 60 views | 30 views | | **LACT** | 120Ā° range | 90Ā° range | 60Ā° range | --- ### Citation ```bibtex @article{liu2024imaging, title={Imaging foundation model for universal enhancement of non-ideal measurement ct}, author={Liu, Yuxin and Ge, Rongjun and He, Yuting and Wu, Zhan and Yang, Shangwen and Gao, Yuan and You, Chenyu and Wang, Ge and Chen, Yang and Li, Shuo}, journal={arXiv preprint arXiv:2410.01591}, year={2024} } ``` ### Links - šŸ“Š [Browse sample data](https://huggingface.co/datasets/YutingHe-list/SimNICT/viewer) - šŸ“„ [Download script](https://huggingface.co/datasets/YutingHe-list/SimNICT/resolve/main/download_simnict.py) - šŸ“‘ [Research paper](https://arxiv.org/abs/2410.01591) --- *For questions or collaborations, contact via the [arXiv paper](https://arxiv.org/abs/2410.01591). Contributions welcome!*