CL-Detection2024

挑战的目标是促进侧位X射线图像中通用头颅测量标志检测的发展。具体来说，检测算法有望准确定位侧位X光片中的53个标记点（13个软组织相关标记点、6个牙齿相关标记点、19个颅骨相关标记点、13个颈椎相关标记点、2个校准尺标记点）。任务——这是全景X射线图像中颈椎标志检测的首个挑战。数据集——提供最多样化、最细致的侧位X射线数据集，包括来自4个医疗中心的约700张2D X射线图像。评估——不仅关注检测准确率，还关注检测效率，这与实际临床实践和要求相一致。

The objective of this challenge is to advance the development of general cephalometric landmark detection in lateral cephalometric X-ray images. Specifically, the detection algorithms are expected to accurately localize 53 landmark points from lateral X-ray radiographs, which are categorized into 13 soft tissue-related landmarks, 6 tooth-related landmarks, 19 skull-related landmarks, 13 cervical vertebra-related landmarks, and 2 calibration ruler landmarks. This task constitutes the first challenge dedicated to cervical vertebra landmark detection in panoramic X-ray images. The provided dataset is a highly diverse and meticulously annotated lateral cephalometric X-ray dataset, including approximately 700 2D X-ray images sourced from 4 medical centers. For evaluation, both detection accuracy and efficiency will be taken into consideration, which aligns with real-world clinical practices and requirements.

搜集汇总

背景与挑战

背景概述

CL-Detection2024是一个医学影像数据集，专注于侧位X射线图像中的头影测量标志点检测。该数据集包含约700张来自4个医疗中心的图像，每张图像标注了53个解剖标志点，涵盖软组织、牙齿、颅骨、颈椎和校准尺等类别。其特点是首次针对全景X射线图像的颈椎标志点检测任务，并提供兼顾检测精度和效率的评估指标，旨在为头影测量标志点检测方法提供全面的基准测试。