File size: 1,368 Bytes
395eb5d eb1a518 395eb5d e98e9aa 339c2dc a46ab76 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 |
# Lung Nodule Segmentation Dataset

## Introduction
Welcome to the Lung Nodule Segmentation Dataset repository! This project aims to provide a comprehensive dataset for researchers and developers to build and evaluate machine learning models for lung nodule segmentation. Accurate detection and segmentation of lung nodules are crucial steps in the early diagnosis and treatment of lung cancer.
## Dataset Overview
The dataset consists of high-resolution CT scans with expertly annotated lung nodules. Each scan is provided with corresponding segmentation masks to facilitate the training and evaluation of segmentation models.
### Features
- **High-Resolution CT Scans:** Detailed CT images of the lungs.
- **Expert Annotations:** Accurate segmentation masks created by medical experts.
- **Diverse Samples:** A variety of lung nodules with different sizes and shapes.
### Dataset Description
The dataset includes images and masks of lung CT scans, annotated by medical professionals. The annotations highlight regions of interest for lung nodule detection, providing a valuable resource for training machine learning models.
**Data Statistics**
- Number of Images to train : 239
- Number of Masks: 239
- Number of Images to train : 41
- Number of Masks: 41
- Image Resolution: 416x416 pixels
- File Format: PNG |