Endoscopy Artefact Detection (EAD) Dataset (includes updated 2020 version)

We aim to establish a first large and comprehensive dataset for "Endoscopy artefact detection". The released data is a part of Endoscopy Artefact Detection IEEE ISBI challenge. With the Endoscopy Artefact Detection Challenge (EAD), we aim to identify hindrances like saturations, motion blur, specular reflections, bubbles, imaging artefacts, contrast and instrument using revolutionary techniques in artificial intelligence. This dataset has been updated with additional class instance (blood) and annotation samples. The provided challenge is important for endoscopy video processing task in obtaining more sensible temporal information. The challenge is sub-divided into three tasks: 1) Multi-class artefact detection: Localization of bounding boxes and class labels for 8 artefact classes for given frames 2) Region segmentation: Precise boundary delineation of detected artefacts for 5 artefact classes 3) Detection generalization: Detection performance independent of specific data type and source (test only) Useful tools for this dataset: https://sharibox.github.io/EAD2019/; https://github.com/sharibox/EndoCV2020 Updated version 2020: We have added an updated dataset of EAD2019 (https://data.mendeley.com/datasets/c7fjbxcgj9/2). Please note that several frames have been included in this dataset. If you use this dataset then please cite the listed works below: 1) Ali, S., Zhou, F., Daul, C., Braden, B., Bailey, A., Realdon, S., East, J., Wagni`eres, G., Loschenov, V., Grisan, E., et al., 2019. Endoscopy artifact detection (EAD 2019) challenge dataset. arXiv preprint arXiv:1905.03209 . 2) Ali, S., Zhou, F., Bailey, A., Braden, B., East, J.E., Lu, X., Rittscher, J., 2021. A deep learning framework for quality assessment and restoration in video endoscopy. Medical Image Analysis 68, 101900. doi:https://doi.org/10.1016/j.media.2020.101900. 3) Ali, S., Zhou, F., Braden, B., Bailey, A., Yang, S., Cheng, G., Zhang, P., Li, X., Kayser, M., Soberanis-Mukul, R.D., et al., 2020. An objective comparison of detection and segmentation algorithms for artefacts in clinical endoscopy. Scientific reports 10, 1–15. 4) Ali, S., Dmitrieva, M., Ghatwary, N. et al. Deep learning for detection and segmentation of artefact and disease instances in gastrointestinal endoscopy. Medical Image Analysis, 2021. doi: https://doi.org/10.1016/j.media.2021.102002.

我们旨在构建首个面向内窥镜伪影检测（Endoscopy artefact detection）的大型综合数据集。本次公开的数据属于IEEE ISBI内窥镜伪影检测挑战赛赛道数据集。依托本次内窥镜伪影检测挑战赛（简称EAD），我们旨在借助前沿人工智能技术，识别各类干扰因素，包括过曝、运动模糊、镜面反射、气泡、成像伪影、对比度异常以及内镜器械干扰。本数据集已完成更新，新增了血液（blood）类别样本及标注数据。

本次挑战赛对于内窥镜视频处理任务获取更合理的时序信息具有重要意义。挑战赛共分为三项任务：
1) 多类别伪影检测：对给定帧内的8类伪影进行边界框定位与类别标注；
2) 区域分割：针对5类伪影实现检测目标的精准边界勾勒；
3) 检测泛化：检测性能不依赖特定数据类型与来源（仅用于测试）。

本数据集配套工具资源：https://sharibox.github.io/EAD2019/；https://github.com/sharibox/EndoCV2020

2020更新版本：我们推出了EAD2019的升级版数据集（https://data.mendeley.com/datasets/c7fjbxcgj9/2），本数据集新增了若干帧图像。

若您使用本数据集，请引用如下文献：
1) Ali S, Zhou F, Daul C, 等. 内窥镜伪影检测（EAD 2019）挑战赛数据集[EB/OL]. arXiv预印本, arXiv:1905.03209, 2019.
2) Ali S, Zhou F, Bailey A, 等. 视频内窥镜质量评估与修复的深度学习框架[J]. 医学图像分析, 2021, 68: 101900. DOI: https://doi.org/10.1016/j.media.2020.101900.
3) Ali S, Zhou F, Braden B, 等. 临床内窥镜伪影检测与分割算法的客观对比[J]. 科学报告, 2020, 10: 1–15.
4) Ali S, Dmitrieva M, Ghatwary N, 等. 胃肠内窥镜伪影与病变实例检测及分割的深度学习方法[J]. 医学图像分析, 2021. DOI: https://doi.org/10.1016/j.media.2021.102002.

搜集汇总

背景与挑战

背景概述

Endoscopy Artefact Detection (EAD) Dataset是一个用于内窥镜伪影检测的大型数据集，包含2020年更新的版本，支持多类伪影检测、区域分割和检测泛化三个任务。该数据集旨在通过人工智能技术识别内窥镜视频中的各种伪影和障碍物，适用于计算机视觉和生物医学成像等领域的研究。