特发性肺纤维化肺部蜂窝样改变影像标注高质量数据集

基于 DCMM 数据成熟度模型规范，通过多来源授权采集，整合多家合作医疗机构的异构数据，在院内指定安全环境中，以合规接口对接、批量导出等方式，采集汇聚医技要素调度数据与肺部蜂窝样改变影像数据。数据经严格去标识化脱敏，剔除敏感信息，我司获医院正式授权拥有运营权。数据仅限合规流通，服务数据要素市场化，全流程遵循法规协议，保障数据安全。Based on the DCMM Data Maturity Model specification, authorized multi-source data collection is carried out. Heterogeneous data from multiple cooperating medical institutions is integrated, and medical technical element scheduling data as well as imaging data of pulmonary honeycomb-like changes are collected and aggregated in the hospital's designated secure environment via compliant interface docking, batch export and other means. The collected data has undergone strict de-identification and desensitization to eliminate sensitive information, and our company has obtained official authorization from the hospital for operational rights. The data is only permitted for compliant circulation to support the marketization of data elements, with the entire process complying with relevant laws, regulations and agreements to guarantee data security.

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于特发性肺纤维化肺部蜂窝样改变的高质量影像标注数据集，数据规模达10TB，通过协议从多家医疗机构整合而来，经过严格去标识化脱敏处理。它主要用于AI预训练场景，支持数据要素市场化合规流通，全流程遵循法规协议，旨在为医疗AI模型开发提供专业、安全的影像数据支持。