通过法国国家医院出院数据库识别慢性血栓栓塞性肺动脉高压

数据集获取（Figshare）

• 源站标题：Identifying chronic thromboembolic pulmonary hypertension through the French national hospital discharge database
• DOI：10.1371/journal.pone.0214649
• 发布日期：2019-04-18
• 文件数：4 个
• 文件总体积：1.44 MB
• 源站页面：https://plos.figshare.com/articles/dataset/Identifying_chronic_thromboembolic_pulmonary_hypertension_through_the_French_national_hospital_discharge_database/8012219

命令行下载

curl -L -o "pone.0214649.s001.tif" "https://ndownloader.figshare.com/files/14926424"
curl -L -o "pone.0214649.s002.docx" "https://ndownloader.figshare.com/files/14926430"
curl -L -o "pone.0214649.s003.docx" "https://ndownloader.figshare.com/files/14926433"
curl -L -o "pone.0214649.s004.docx" "https://ndownloader.figshare.com/files/14926439"

文件清单

• pone.0214649.s001.tif（1.39 MB）
• pone.0214649.s002.docx（12.69 KB）
• pone.0214649.s003.docx（17.26 KB）
• pone.0214649.s004.docx（13.29 KB）

文件总体积 1.44 MB（源站 API 实测）。

数据加载示例（图像类）

from PIL import Image
import glob, os

files = (glob.glob(os.path.join(path, "**", "*.png"), recursive=True)
       + glob.glob(os.path.join(path, "**", "*.jpg"), recursive=True)
       + glob.glob(os.path.join(path, "**", "*.tif"), recursive=True))
print("图像文件数:", len(files))
img = Image.open(files[0]); print("尺寸/模式:", img.size, img.mode)

# torchvision Dataset 方式:
# from torchvision import datasets
# ds = datasets.ImageFolder(path)  # 要求 子目录=类别

目录组织与标注格式以源站说明和下载后实际文件为准。

数据说明

Chronic thromboembolic pulmonary hypertension (CTEPH), a rare pulmonary vascular disease, is often misdiagnosed due to nonspecific symptoms. The objective of the study was to develop, refine and validate a case ascertainment algorithm to identify CTEPH patients within the French exhaustive hospital discharge database (PMSI), and to use it to estimate the annual number of hospitalized patients with CTEPH in France in 2015, as a proxy for disease prevalence. As ICD-10 coding specifically for CTEPH was not available at the time of the study, a case ascertainment algorithm was developed in close collaboration with an expert committee, using a two-step process (refinement and validation), based on matched data from PMSI and hospital medical records from 2 centres. The best-performing algorithm (specificity 95%, sensitivity 70%) consisted of ≥1 pulmonary hypertension (PH) diagnosis during 2015 and any of the following criteria over 2009–2015: (i) CTEPH interventional procedure, (ii) admission for PH and pulmonary embolism (PE), (iii) PE followed by hospitalization in competence centre then in reference centre, (iv) history of PE and right heart catheterization. Patients with conditions suggestive of pulmonary arterial hypertension were excluded. A total of 3,138 patients hospitalized for CTEPH was estimated for 2015 (47 cases/million, range 43 to 50 cases/million). Assuming that patients are hospitalized at least once a year, the present study provides an estimate of the minimal prevalence of CTEPH and confirms the heavy burden of this disease.

许可

CC BY 4.0

引用

学术使用请引用 DOI 10.1371/journal.pone.0214649（source: https://plos.figshare.com/articles/dataset/Identifying_chronic_thromboembolic_pulmonary_hypertension_through_the_French_national_hospital_discharge_database/8012219）。