OCTA-500

Optical coherence tomography angiography (OCTA) is a novel imaging modality that allows a micron-level resolution to present the three-dimensional structure of the retinal vascular.We propose a new multi-modality dataset, dubbed OCTA-500. It contains 500 subjects with 2 field of view (FOV) types, including OCT and OCTA volumes, 6 types of projections, 4 types of text labels and 2 types of pixel-level labels. This dataset contains more than 360K images with a size of about 80GB. Now, OCTA-500 is publicly available. This dataset is released for academic research use only.OCTA-500 includes two subsets: OCTA_6M and OCTA_3M. OCTA_6M(No.10001-No.10300):FOV: 6mm*6mm*2mmVolume: 400pixel*400pixel*640pixelOCTA_3M(No.10301-No.10500):FOV: 3mm*3mm*2mmVolume: 304pixel*304pixel*640pixel Both subsets contain the following information:OCT volumesOCTA volumesProjection Maps-OCT FULL(average)-OCT ILM_OPL (average)-OCT OPL_BM (average)-OCTA FULL (average)-OCTA ILM_OPL (maximum)-OCTA OPL_BM (maximum)Text Label-Gender-Age-OS/OD-DiseasePixel Label-retinal vessel segmentation-foveal avascular zone segmentation By using the OCTA-500 dataset, you are obliged to reference at least one of the following papers:-Mingchao Li, Yerui Chen, Zexuan Ji, Keren Xie, Songtao Yuan, Qiang Chen, and Shuo Li.“Image projection network: 3D to 2D image segmentation in OCTA images,” IEEE Trans. Med. Imaging, vol. 39, no. 11 pp. 3343-3354, 2020.-Mingchao Li, Yuhan Zhang, Zexuan Ji, Keren Xie, Songtao Yuan, Qinghuai Liu and Qiang Chen. "IPN-V2 and OCTA-500: Methodology and Dataset for Retinal Image Segmentation," arXiv:2012.07261. Related Codes:IPN: https://github.com/chaosallen/IPN_tensorflowIPN-V2: https://github.com/chaosallen/IPNV2_pytorch Get Password:To get the password of the compressed package, an application email must be sent to [e-mail] with a specified form like below, otherwise may be ignored.Title of Mail:The string of 'OCTA500' can not be empty. It is the fixed form and a special sign we use to identifying your downloading intention from other disturbers like spams.The contents appending to OCTA500 can help us identifying you more easily.OCTA500: your_organization: your_nameBody of Mail:Organization Detail: Your Organization DetailsMain Works: Your Main Works Usages: YourUsages About This Data Set

光学相干断层扫描血管造影（Optical coherence tomography angiography, OCTA）是一种新型成像技术，可实现微米级分辨率，呈现视网膜血管的三维结构。我们提出了一款全新的多模态数据集，命名为OCTA-500。该数据集涵盖500名受试者，包含2种视野（field of view, FOV）类型，同时收录了OCT与OCTA容积数据、6类投影图像、4类文本标签以及2类像素级标签。数据集包含超过36万张图像，总数据量约80GB。目前OCTA-500已公开上线，仅用于学术研究用途。
OCTA-500包含两个子数据集：OCTA_6M与OCTA_3M。其中OCTA_6M（编号10001至10300）：FOV为6mm×6mm×2mm，体素尺寸为400像素×400像素×640像素；OCTA_3M（编号10301至10500）：FOV为3mm×3mm×2mm，体素尺寸为304像素×304像素×640像素。
两个子数据集均包含以下内容：
1. OCT容积数据；
2. OCTA容积数据；
3. 投影图像：
- OCT全图（平均投影）
- OCT内界膜（Internal Limiting Membrane, ILM）-外丛状层（Outer Plexiform Layer, OPL）（平均投影）
- OCT外丛状层（Outer Plexiform Layer, OPL）-Bruch膜（Bruch's Membrane, BM）（平均投影）
- OCTA全图（平均投影）
- OCTA内界膜（Internal Limiting Membrane, ILM）-外丛状层（Outer Plexiform Layer, OPL）（最大投影）
- OCTA外丛状层（Outer Plexiform Layer, OPL）-Bruch膜（Bruch's Membrane, BM）（最大投影）
4. 文本标签：
- 性别
- 年龄
- 眼别（oculus sinister, OS；oculus dextr, OD，即左眼/右眼）
- 疾病标签
5. 像素级标签：
- 视网膜血管分割标签
- 中心凹无血管区分割标签

使用OCTA-500数据集时，需至少引用以下论文之一：
1. Mingchao Li, Yerui Chen, Zexuan Ji, Keren Xie, Songtao Yuan, Qiang Chen, and Shuo Li. "Image projection network: 3D to 2D image segmentation in OCTA images," *IEEE Transactions on Medical Imaging*, vol. 39, no. 11, pp. 3343-3354, 2020.
2. Mingchao Li, Yuhan Zhang, Zexuan Ji, Keren Xie, Songtao Yuan, Qinghuai Liu and Qiang Chen. "IPN-V2 and OCTA-500: Methodology and Dataset for Retinal Image Segmentation," arXiv:2012.07261.

相关代码：
IPN：https://github.com/chaosallen/IPN_tensorflow
IPN-V2：https://github.com/chaosallen/IPNV2_pytorch

获取压缩包密码：
如需获取压缩包密码，需向[e-mail]发送指定格式的申请邮件，否则申请将被忽略。
邮件主题需包含字符串"OCTA500"（不可为空，这是我们用于区分垃圾邮件等干扰项、识别下载意向的固定标识与特殊标记），可在"OCTA500"后补充相关信息以便我们更精准地识别申请人。
邮件格式示例：
主题：OCTA500: your_organization: your_name
邮件正文：
Organization Detail: 您的机构详情
Main Works: 您的主要研究成果
Usages: 您使用本数据集的用途