窄带成像内镜评估声带病变的学习曲线数据集

数据集获取（Figshare）

• 源站标题：Learning curve for endoscopic evaluation of vocal folds lesions with narrow band imaging
• DOI：10.6084/m9.figshare.11390991.v1
• 发布日期：2019-12-18
• 文件数：10 个
• 文件总体积：2.55 MB
• 源站页面：https://scielo.figshare.com/articles/dataset/Learning_curve_for_endoscopic_evaluation_of_vocal_folds_lesions_with_narrow_band_imaging/11390991

命令行下载

curl -L -o "dc6d8067785e1525a0d11eb4c35735a565c7ca82.jpg" "https://ndownloader.figshare.com/files/35672345"
curl -L -o "0c9222226af1dd759db6acec28b6d12acefe7230.jpg" "https://ndownloader.figshare.com/files/35672348"
curl -L -o "11b89fef42685e842b5ab85c2ede09028ab4a7d2.jpg" "https://ndownloader.figshare.com/files/35672351"
curl -L -o "15c72b1ecfba0c3fa6cbca505f089197aa5fbfb6.jpg" "https://ndownloader.figshare.com/files/35672354"
curl -L -o "45d8a4384df92019104af8eb8dab31e17206b362.jpg" "https://ndownloader.figshare.com/files/35672357"
curl -L -o "06f991a0f10c9b307d190691da6dcf8cfb726534.jpg" "https://ndownloader.figshare.com/files/35672360"
curl -L -o "adb18d910799255466fc0ae1068a657640d87a38.jpg" "https://ndownloader.figshare.com/files/35672363"
curl -L -o "a0047c57ae821adb788add502fdf594591e8a58c.jpg" "https://ndownloader.figshare.com/files/35672366"

文件清单

• dc6d8067785e1525a0d11eb4c35735a565c7ca82.jpg（197.93 KB）
• 0c9222226af1dd759db6acec28b6d12acefe7230.jpg（159.24 KB）
• 11b89fef42685e842b5ab85c2ede09028ab4a7d2.jpg（391.67 KB）
• 15c72b1ecfba0c3fa6cbca505f089197aa5fbfb6.jpg（316.46 KB）
• 45d8a4384df92019104af8eb8dab31e17206b362.jpg（194.62 KB）
• 06f991a0f10c9b307d190691da6dcf8cfb726534.jpg（209.77 KB）
• adb18d910799255466fc0ae1068a657640d87a38.jpg（164.32 KB）
• a0047c57ae821adb788add502fdf594591e8a58c.jpg（394.76 KB）
• 6d5097e6ed89f2368bb347f98b49a97a04bdfb03.jpg（323.21 KB）
• 4056dff731922a770b248fc2d92df967e9c78ba7.jpg（198.25 KB）

文件总体积 2.55 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

Abstract Introduction: The endoscopic methods are progressing and becoming more common in routine clinical diagnosis in the field of otorhinolaryngology. Relatively large amount of researches have proved high accuracy of narrow band imaging endoscopy in differentiating benign and malignant lesions within vocal folds. However, little is known about learning curve in narrow band imaging evaluation of laryngeal lesions. Objective: The aim of this study was to determine the learning curve for the narrow band imaging evaluation of vocal folds pathologies depending on the duration of the procedure. Methods: Records of 134 narrow band imaging that were analyzed in terms of the duration of the procedure and the accuracy of diagnosis confirmed by histopathological diagnosis were enrolled in the study. The narrow band imaging examinations were performed sequentially by one investigator over a period of 18 months. Results: The average duration of narrow band imaging recordings was 127.82 s. All 134 studies were divided into subsequent series of several elements. An evident decrease in time of investigation was noticed between 13th and 14th series, when the examinations were divided into 5 elements series, which corresponds to the difference between 65th and 70th subsequent narrow band imaging examination. Parallel groups of 67 examinations were created. Group 1 included 1st to 67th subsequent narrow band imaging examination; Group 2 – 68th to 134th narrow band imaging examinations. The non-parametric U Mann–Whitney test confirmed statistically significant difference between the mean duration of narrow band imaging examination in both groups 160.5 s and 95.1 s, respectively (p < 10−7). Sensitivity and specificity of narrow band imaging examination in the first group were respectively: 83.7% and 76.7%. In the second group, these indicators amounted 98.1% and 80% respectively. Conclusions: A minimum of 65th–70th narrow band imaging examinations are required to reach a plateau phase of the learning process in assessment of glottis lesions. Analysis of learning curves is useful for the development of training programs and determination of a mastery level.

许可

CC BY 4.0

引用

学术使用请引用 DOI 10.6084/m9.figshare.11390991.v1（source: https://scielo.figshare.com/articles/dataset/Learning_curve_for_endoscopic_evaluation_of_vocal_folds_lesions_with_narrow_band_imaging/11390991）。