CBCTLiTS
收藏资源简介:
CBCTLiTS数据集由萨尔茨堡应用科学大学和萨尔茨堡大学创建,是一个合成生成的CBCT/CT配对数据集,包含201个样本,用于医学图像分割和风格转换研究。数据集包含不同质量级别的CBCT图像,从高质量到严重伪影,支持多种研究场景如单模态和多模态分割、多任务学习和风格转换。数据集的创建基于LiTS数据集,通过模拟CBCT扫描和CT图像的对齐,生成合成数据。该数据集主要应用于计算机辅助干预和放射学领域,旨在提高CBCT图像分析的准确性和效率。
The CBCTLiTS dataset, created by Salzburg University of Applied Sciences and University of Salzburg, is a synthetically generated paired CBCT/CT dataset containing 201 samples, intended for medical image segmentation and style transfer research. It includes CBCT images with varying quality levels, ranging from high-quality scans to those with severe artifacts, supporting multiple research scenarios such as unimodal and multimodal segmentation, multi-task learning, and style transfer. The dataset is developed based on the LiTS dataset, with synthetic data generated by simulating the alignment between CBCT scans and CT images. It is primarily applied in the fields of computer-assisted intervention and radiology, aiming to improve the accuracy and efficiency of CBCT image analysis.




