Endotect 2021

在人类一生中，人类消化系统容易遭受许多不同的疾病和异常。其中一些可能危及生命，并对患者的健康和福祉构成严重风险。在大多数情况下，如果足够早地进行致命疾病的检测，则可以以很高的机会完全治愈。因此，重要的是在胃肠道的常规检查中识别并报告所有病变。目前，进行这些调查的黄金标准是通过视频内窥镜检查，这是一种涉及将小型摄像机连接到经口或直肠插入的管子上的程序。然而，这个程序有一个主要的缺点。该方法高度依赖于操作内窥镜的人的技能和经验，这反过来导致高操作者变化和性能。这是在测量息肉检测性能时高未命中率的原因之一，其中一些未命中率高达20%。我们认为这是通过在内窥镜检查期间实时进行的自动框架分析来帮助医生检测病变的机会。模式识别社区拥有丰富的知识，可以帮助完成此任务，使其非常适合ICPR。在这场比赛中所做的工作有可能产生真正的社会影响，因为它直接影响卫生保健专业人员可以提供的护理质量。

Over the course of a human lifetime, the digestive system is susceptible to a wide range of diseases and abnormalities. Some of these can be life-threatening, posing severe risks to patients' health and well-being. In most cases, if life-threatening diseases are detected sufficiently early, they can be fully cured with a high success rate. Therefore, it is critical to identify and report all lesions during routine gastrointestinal examinations. Currently, the gold standard for these investigations is video endoscopy, a procedure that involves attaching a small camera to a tube inserted via the mouth or rectum. However, this method has a major drawback: it is highly dependent on the skill and experience of the endoscopist performing the procedure, leading to significant inter-operator variability and inconsistent performance. This is one of the key reasons for the high miss rate in polyp detection assessments, with some miss rates reaching as high as 20%. We see this as an opportunity to assist clinicians in detecting lesions through real-time automated frame analysis during endoscopy. The pattern recognition community has accumulated extensive knowledge applicable to this task, making it highly suitable for ICPR. The work of this competition has the potential to create genuine social impact, as it directly affects the quality of care that healthcare professionals can provide.

搜集汇总

背景与挑战

背景概述

Endotect 2021是一个专注于消化系统疾病检测的数据集，旨在通过内窥镜视频的实时自动分析帮助医生识别病变，由挪威多所研究机构联合发布，以提高医疗检测的准确性和效率。