意大利癌症登记质量检查系统比较数据集

数据集获取（Figshare）

• 源站标题：Table_1_Comparison between two cancer registry quality check systems: functional features and differences in an Italian network of cancer registries dataset.docx
• DOI：10.3389/fonc.2023.1197942.s001
• 发布日期：2023-05-25
• 文件数：1 个
• 文件总体积：699.38 KB
• 源站页面：https://frontiersin.figshare.com/articles/dataset/Table_1_Comparison_between_two_cancer_registry_quality_check_systems_functional_features_and_differences_in_an_Italian_network_of_cancer_registries_dataset_docx/23199551

命令行下载

curl -L -o "Table_1_Comparison between two cancer registry quality check systems: functional features and differences in an Italian network of cancer registries d.docx" "https://ndownloader.figshare.com/files/40895780"

文件清单

• Table_1_Comparison between two cancer registry quality check systems: functional features and differences in an Italian network of cancer registries d.docx（699.38 KB）

文件总体积 699.38 KB（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

PurposeThe aim of this study was to compare the functional characteristics of two computer-based systems for quality control of cancer registry data through analysis of their output differences.
MethodsThe study used cancer incidence data from 22 of the 49 registries of the Italian Network of Cancer Registries registered between 1986 and 2017. Two different data checking systems developed by the WHO International Agency for Research on Cancer (IARC) and the Joint Research Center (JRC) with the European Network of Cancer Registries (ENCR) and routinely used by registrars were used to check the quality of the data. The outputs generated by the two systems on the same dataset of each registry were analyzed and compared.
ResultsThe study included a total of 1,305,689 cancer cases. The overall quality of the dataset was high, with 86% (81.7-94.1) microscopically verified cases and only 1.3% (0.03-3.06) cases with a diagnosis by death certificate only. The two check systems identified a low percentage of errors (JRC-ENCR 0.17% and IARC 0.003%) and about the same proportion of warnings (JRC-ENCR 2.79% and IARC 2.42%) in the dataset. Forty-two cases (2% of errors) and 7067 cases (11.5% of warnings) were identified by both systems in equivalent categories. 11.7% of warnings related to TNM staging were identified by the JRC-ENCR system only. The IARC system identified mainly incorrect combination of tumor grade and morphology (72.5% of warnings).
ConclusionBoth systems apply checks on a common set of variables, but some variables are checked by only one of the systems (for example, checks on patient follow-up and tumor stage at diagnosis are included by the JRC-ENCR system only). Most errors and warnings were categorized differently by the two systems, but usually described the same issues, with warnings related to “morphology” (JRC-ENCR) and “histology” (IARC) being the most frequent. It is important to find the right balance between the need to maintain high standards of data quality and the workability of such systems in the daily routine of the cancer registry.

许可

CC BY 4.0

引用

学术使用请引用 DOI 10.3389/fonc.2023.1197942.s001（source: https://frontiersin.figshare.com/articles/dataset/Table_1_Comparison_between_two_cancer_registry_quality_check_systems_functional_features_and_differences_in_an_Italian_network_of_cancer_registries_dataset_docx/23199551）。