成人甲状腺结节超声影像属性标注数据集

本数据集为合肥合滨智能机器人有限公司自行生产的甲状腺结节属性标注数据集，原始数据采集及后续加工处理均由本单位全程主导实施，合肥合滨智能机器人有限公司向合肥巢电科技有限公司采购视频采集相关软硬件设备，完成设备调试后在合作医院部署。采集完成的原始超声图像由合肥合滨智能机器人有限公司技术团队闭环带回公司，通过自主研发的超声影像数据预处理工具，完成图片处理，再采用《中国甲状腺影像报告和数据系统（CTIRADS）2020 版》，由具备3年以上甲状腺影像标注经验的团队完成结节方位、边缘、回声等信息的人工标注及主治医师及以上职称人员交叉审核，确保标注准确性与规范性，最终形成结构化甲状腺结节属性标注数据集。This dataset is a thyroid nodule attribute annotation dataset independently developed and produced by Hefei Hebin Intelligent Robotics Co., Ltd. The entire process of original data collection and subsequent processing was fully led and implemented by our company. Hefei Hebin Intelligent Robotics Co., Ltd. purchased hardware and software equipment related to video collection from Hefei Chaodian Technology Co., Ltd., and deployed the equipment in cooperative hospitals after completing debugging. The collected original ultrasound images were brought back to the company in a closed-loop manner by the technical team of Hefei Hebin Intelligent Robotics Co., Ltd., and processed using independently developed ultrasound image data preprocessing tools. Then, following the guidelines of *Chinese Thyroid Imaging Reporting and Data System (CTIRADS) 2020 Edition*, a team with more than 3 years of experience in thyroid image annotation completed manual annotation of information such as nodule orientation, margin, and echo, followed by cross-review by attending physicians and personnel with equivalent or higher professional titles to ensure the accuracy and standardization of annotations. Ultimately, a structured thyroid nodule attribute annotation dataset was formed.

搜集汇总

背景与挑战

背景概述

该数据集包含600条成人甲状腺结节超声影像的标注数据，每月更新，每条数据均提供详细的结节属性标注，如方位、边缘、回声等，严格遵循CTIRADS 2020标准。它适用于医院超声科和外科医师辅助结节风险评估与诊疗，以及医学院校的教学培训，旨在提升甲状腺结节的诊断准确性和临床教育效果。