妇科宫颈癌病理数字切片标注标准数据集

计算机以特定方式存储与组织数据。本系统采用混合存储架构：图像以病理全数字切片（WSI）格式按切片单元存储；文字按病例单元结构化存储；标注信息则以JSON格式独立存储。所有数据部署于加密数据库，通过权限矩阵实现分级访问控制，并支持自动化安全审计及基于哈希索引的快速检索。
在字段设置上，包含多个关键字段及其属性：病例ID，作为病例的唯一标识符，字符型，便于数据追踪与管理；图像名，病理全数字切片的唯一标识，字符型，与病例ID建立映射；所属医院，样本来源机构，字符型，采用标准化编码；样本部位，解剖学位置，字符型；WHO分类，疾病分类，字符型，对应WHO编码；临床诊断，临床诊断信息，文本型，结构化描述；临床资料，病史数据，文本型；标注信息，肿瘤与正常组织标注坐标，JSON格式，含边界框及像素级数据；图像评价，病理专家评估结果，字符型，三级评分。
主关键字段包括组别、性别、年龄、样本类型、HPV检测结果、TCT检测结果。通过病例ID、图像名及标注信息字段的组合，确保每条数据记录的唯一性。

Computers store and organize data in specific manners. This system adopts a hybrid storage architecture: images are stored in slice units in the format of Whole Slide Images (WSI); textual data is structurally stored in case units; annotation information is independently stored in JSON format. All data is deployed in an encrypted database, where hierarchical access control is implemented through a permission matrix, and automated security audit as well as fast retrieval based on hash indexes are supported.
Regarding field settings, multiple key fields and their attributes are included: Case ID, the unique identifier of a case, character type, which facilitates data tracking and management; Image name, the unique identifier of Whole Slide Images, character type, which establishes a mapping relationship with Case ID; Affiliated Hospital, the source institution of the sample, character type, adopting standardized coding; Sample Site, anatomical location, character type; WHO Classification, disease classification, character type, corresponding to WHO codes; Clinical Diagnosis, clinical diagnosis information, text type with structured description; Clinical Data, medical history data, text type; Annotation Information, annotation coordinates of tumor and normal tissues, in JSON format, including bounding boxes and pixel-level data; Image Evaluation, pathological expert assessment results, character type with three-level scoring.
The core key fields include group, gender, age, sample type, HPV test result and TCT test result. The combination of Case ID, Image Name and Annotation Information fields ensures the uniqueness of each data record.

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于妇科宫颈癌病理数字切片的标注标准资源，旨在为医学研究和诊断提供规范化的图像标注数据。它可能包含经过专业标注的病理切片图像，用于支持宫颈癌的检测、分类或算法开发，并作为数据加工集合的一部分，体现了在医疗数据标准化方面的应用价值。