OCT B-scan图像中央凹无血管区面积估计数据集

数据集获取（Figshare）

• 源站标题：dataset
• DOI：10.6084/m9.figshare.27262257.v1
• 发布日期：2024-10-24
• 文件数：3 个
• 文件总体积：22.79 MB
• 源站页面：https://figshare.com/articles/dataset/dataset/27262257

命令行下载

curl -L -o "model_1_dataset.zip" "https://ndownloader.figshare.com/files/49880379"
curl -L -o "model_2_dataset.csv" "https://ndownloader.figshare.com/files/49880382"
curl -L -o "model3_dataset.csv" "https://ndownloader.figshare.com/files/49880385"

文件清单

• model_1_dataset.zip（22.77 MB）
• model_2_dataset.csv（9.43 KB）
• model3_dataset.csv（8.36 KB）

文件总体积 22.79 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

This repository contains the datasets used for training and evaluating Models 1, 2, and 3 in our study on estimating the Foveal Avascular Zone (FAZ) area from OCT B-scan images using machine learning algorithms.

Model 1 Dataset
Objective
To predict the FAZ length from OCT B-scan images using machine learning.
Data Details
Images
Total of 1,146 OCT B-scan images.
Image size: 128 × 128 pixels.
Each image is a B-scan slice centered on the fovea.
Images are named sequentially (e.g., “1.bmp”, “2.bmp”, “3.bmp”), with no patient identification information included.
Labels
Each image is associated with a numerical value representing the horizontal length of the FAZ segment.
The length corresponds to the width of the FAZ at the foveal center.
Length values are provided in pixels.
Files Included
B-scan Images
Provided in BMP format.
Filenames are numbered sequentially to maintain anonymity.
Labels File
“model_1_dataset.csv” containing two columns:
“file_name”: Name of the image file.
“length”: Corresponding FAZ length in pixels.

Model 2 Dataset
Objective
To predict the FAZ area in pixels from length measurements using machine learning.
Data Details
Length Measurements
Lengths are measured across up to five horizontal lines through the fovea.
Lengths are labeled as “length1” to “length5”:
“length3”: Central line passing through the foveal center.
“length2” and “length4”: Lines above and below the central line.
“length1” and “length5”: Lines further above and below the central line.
Each line represents a different horizontal slice of the OCT image.
Labels
Each set of length measurements corresponds to a FAZ area measured in pixels (“Faz_pixel”).
Model Variations
Model 2 (1 Line)
Uses only the central length (“length3”) for prediction.
Model 2 (3 Lines)
Uses “length2”, “length3”, and “length4” (central line and one line above and below).
Model 2 (5 Lines)
Uses “length1” to “length5” (central line and two lines above and below).
Files Included
Length Measurements and Labels File
“model_2_dataset.csv” containing:
“length1” to “length5”: Length measurements in pixels.
“Faz_pixel”: Corresponding FAZ area in pixels.

Model 3 Dataset
Objective
To convert the FAZ area from pixel

许可

CC BY 4.0

引用

学术使用请引用 DOI 10.6084/m9.figshare.27262257.v1（source: https://figshare.com/articles/dataset/dataset/27262257）。