harvardairobotics/Harvard-GDP

---
license: cc-by-nc-nd-4.0
task_categories:
- image-classification
modality:
- image
language:
- en
tags:
- medical
- ophthalmology
- fairness
- glaucoma
- OCT
- progression
- semi-supervised
pretty_name: Harvard-GDP
size_categories:
- n<1K
---

# Dataset Card: Harvard-GDP

## Dataset Summary

Harvard-GDP (Harvard Glaucoma Detection and Progression) is a multimodal multitask ophthalmology dataset for glaucoma detection and progression forecasting. It is the **largest publicly available glaucoma detection dataset with 3D OCT imaging data** and the **first publicly available glaucoma progression forecasting dataset**. The dataset includes detailed demographic annotations (sex, race) to support fairness learning research.

This dataset was introduced at ICCV 2023: [Harvard Glaucoma Detection and Progression: A Multimodal Multitask Dataset and Generalization-Reinforced Semi-Supervised Learning](https://openaccess.thecvf.com/content/ICCV2023/papers/Luo_Harvard_Glaucoma_Detection_and_Progression_A_Multimodal_Multitask_Dataset_and_ICCV_2023_paper.pdf).

## Dataset Details

### Dataset Description

| Field | Value |
|----------------|-------|
| **Institution**| Department of Ophthalmology, Harvard Medical School |
| **Tasks** | Glaucoma detection, glaucoma progression forecasting |
| **Modality** | OCT RNFLT maps, visual field (MD, TDs) |
| **Scale** | 1,000 patients, 1,000 OCT RNFLT maps |
| **Image size** | 225 × 225 (RNFLT map) |
| **License** | [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/) |

- **Curated by:** Yan Luo, Min Shi, Yu Tian, Tobias Elze, Mengyu Wang
- **License:** [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/) — non-commercial research only
- **Paper:** [ICCV 2023](https://openaccess.thecvf.com/content/ICCV2023/papers/Luo_Harvard_Glaucoma_Detection_and_Progression_A_Multimodal_Multitask_Dataset_and_ICCV_2023_paper.pdf)
- **Contact:** [e-mail], [e-mail]

### Data Fields

Each subject is stored as a `.npz` file (`data_0001.npz` … `data_1000.npz`) in the `rnflt_maps/` folder:

| Field | Description |
|---------------|-------------|
| `rnflt` | OCT retinal nerve fiber layer thickness (RNFLT) map, size 225 × 225 |
| `glaucoma` | Glaucomatous status: `0` = non-glaucoma, `1` = glaucoma |
| `progression` | Vector of 6 progression labels: `0` = non-progression, `1` = progression (first 500 subjects only) |
| `md` | Mean deviation value of visual field |
| `tds` | 52 total deviation values of visual field |
| `age` | Patient age |
| `male` | Gender: `0` = Female, `1` = Male |
| `race` | Patient race |
| `hispanic` | Patient ethnicity |

### Progression Labels

Progression labels are defined for the first 500 subjects (`data_0001` – `data_0500`) only, using 6 criteria:

| Index | Criterion |
|-------|-----------|
| `progression[0]` | MD-based |
| `progression[1]` | VFI-based |
| `progression[2]` | TD pointwise |
| `progression[3]` | MD fast |
| `progression[4]` | MD fast (no p-value cutoff) |
| `progression[5]` | TD pointwise (no p-value cutoff) |

## Uses

### Direct Use

- Glaucoma detection benchmarking with 3D OCT imaging data
- Glaucoma progression forecasting (unimodal and multimodal)
- Semi-supervised learning research with limited labeled medical data
- Fairness and demographic disparity analysis in ophthalmic AI

### Out-of-Scope Use

Clinical decisions, patient care, or any commercial application. This dataset shall not be used for clinical decisions at any time.

## Access

The "Harvard" designation indicates the dataset originates from the Department of Ophthalmology at Harvard Medical School. It does not imply endorsement, sponsorship, or legal responsibility by Harvard University or Harvard Medical School.

## Citation

**BibTeX:**

```bibtex
@inproceedings{luo2023harvard,
title={Harvard Glaucoma Detection and Progression: A Multimodal Multitask Dataset and Generalization-Reinforced Semi-Supervised Learning},
author={Luo, Yan and Shi, Min and Tian, Yu and Elze, Tobias and Wang, Mengyu},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={20471--20482},
year={2023}
}
```

**APA:**

Luo, Y., Shi, M., Tian, Y., Elze, T., & Wang, M. (2023). Harvard Glaucoma Detection and Progression: A Multimodal Multitask Dataset and Generalization-Reinforced Semi-Supervised Learning. *Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV 2023)*, 20471–20482.