广东省某医院体检科的体检数据

体检数据集为广东省某三甲医院长达八年的体检数据记录。在论文中，我们基于此大规模真实世界数据，探索了基于体重指数、血压、性别、血糖和血脂的骨密度筛查策略。相较于传统的基于X-rays的骨密度筛查策略，基于体检数据的筛查法适用范围更广，对身体的累计伤害更小。数据集主要记录了体检人身高、体重、血压、性别、血糖和血脂等体检指标。主要包括1个文件，数据量6.4GB。由于数据的高度私密性，我们随机抽取了1万条数据用作汇交，主要包括1个文件，数据量3514KB。

This physical examination dataset contains 8-year-long physical examination records from a top-tier tertiary hospital in Guangdong Province. In this paper, we explored bone mineral density (BMD) screening strategies based on routine physical examination indicators including body mass index (BMI), blood pressure, sex, blood glucose and blood lipids using this large-scale real-world dataset. Compared with the traditional X-ray-based BMD screening strategy, the screening method based on physical examination data features a wider application range and causes less cumulative harm to the human body. The dataset mainly records physical examination indicators such as height, weight, blood pressure, sex, blood glucose and blood lipids of the examinees. The full dataset primarily includes one file with a total size of 6.4 GB. Owing to the high privacy sensitivity of the original data, we randomly selected 10,000 records for public submission, which also consists of one file with a size of 3514 KB.

搜集汇总

背景与挑战

背景概述

该数据集是广东省某三甲医院长达八年的体检数据记录，包含身高、体重、血压、性别、血糖和血脂等关键指标，用于研究基于体检数据的骨密度筛查策略，以替代传统X射线方法，具有更广适用范围和更低身体伤害。数据集原始规模为6.4GB，但汇交版本为随机抽取的1万条数据（3.46MB），体现了对隐私保护的处理，并隶属于人工智能学科，支持相关医疗研究。