Suppr超能文献

基于几何特征的超广角荧光素血管造影图像亮度补偿。

Ultra-Widefield Fluorescein Angiography Image Brightness Compensation Based on Geometrical Features.

机构信息

Faculty of Biomedical Engineering, Silesian University of Technology, Roosevelta St. 40, 41-800 Zabrze, Poland.

Clinical Department of Ophthalmology, Faculty of Medical Sciences in Zabrze, Medical University of Silesia, Panewnicka St. 65, 40-760 Katowice, Poland.

出版信息

Sensors (Basel). 2021 Dec 21;22(1):12. doi: 10.3390/s22010012.

Abstract

Ultra-widefield fluorescein angiography (UWFA) is an emerging imaging modality used to characterise pathologies in the retinal vasculature, such as microaneurysms (MAs) and vascular leakages. Despite its potential value for diagnosis and disease screening, objective quantitative assessment of retinal pathologies by UWFA is currently limited because laborious manual processing is required. In this report, we describe a geometrical method for uneven brightness compensation inherent to UWFA imaging technique. The correction function is based on the geometrical eyeball shape, therefore it is fully automated and depends only on pixel distance from the center of the imaged retina. The method's performance was assessed on a database containing 256 UWFA images with the use of several image quality measures that show the correction method improves image quality. The method is also compared to the commonly used CLAHE approach and was also employed in a pilot study for vascular segmentation, giving a noticeable improvement in segmentation results. Therefore, the method can be used as an image preprocessing step in retinal UWFA image analysis.

摘要

超广角荧光素血管造影术(UWFA)是一种新兴的成像方式,用于描述视网膜血管的病变,如微动脉瘤(MAs)和血管渗漏。尽管它在诊断和疾病筛查方面具有潜在的价值,但 UWFA 对视网膜病变的客观定量评估目前受到限制,因为需要进行繁琐的手动处理。在本报告中,我们描述了一种针对 UWFA 成像技术固有的不均匀亮度补偿的几何方法。校正函数基于眼球的几何形状,因此它是全自动的,仅取决于与成像视网膜中心的像素距离。该方法在包含 256 张 UWFA 图像的数据库上进行了性能评估,使用了几种图像质量度量标准,这些标准表明校正方法可以提高图像质量。该方法还与常用的 CLAHE 方法进行了比较,并在血管分割的试点研究中得到了应用,分割结果得到了显著改善。因此,该方法可以作为视网膜 UWFA 图像分析中的图像预处理步骤。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/7f74/8747562/0465ea23f90f/sensors-22-00012-g001.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验