Neves João C, Castro Helena, Tomás Ana, Coimbra Miguel, Proença Hugo
Department of Computer Science, IT-Instituto de Telecomunicações, University of Beira Interior, Covilhã, Portugal.
Cytometry A. 2014 Jun;85(6):491-500. doi: 10.1002/cyto.a.22465. Epub 2014 Apr 9.
Life scientists often must count cells in microscopy images, which is a tedious and time-consuming task. Automatic approaches present a solution to this problem. Several procedures have been devised for this task, but the majority suffer from performance degradation in the case of cell overlap. In this article, we propose a method to determine the positions of macrophages and parasites in fluorescence images of Leishmania-infected macrophages. The proposed strategy is primarily based on blob detection, clustering, and separation using concave regions of the cells' contours. In comparison with the approaches of Nogueira (Master's thesis, Department of University of Porto Computer Science, 2011) and Leal et al. (Proceedings of the 9th international conference on Image Analysis and Recognition, Vol. II, ICIAR'12. Berlin, Heidelberg: Springer-Verlag; 2012. pp. 432-439), which also addressed this type of image, we conclude that the proposed methodology achieves better performance in the automatic annotation of Leishmania infections.
生命科学家常常需要对显微镜图像中的细胞进行计数，这是一项繁琐且耗时的任务。自动方法为解决这一问题提供了一种方案。针对此任务已设计出多种程序，但大多数方法在细胞重叠的情况下会出现性能下降的问题。在本文中，我们提出了一种在利什曼原虫感染巨噬细胞的荧光图像中确定巨噬细胞和寄生虫位置的方法。所提出的策略主要基于斑点检测、聚类以及利用细胞轮廓的凹面区域进行分离。与诺盖拉（硕士论文，波尔图大学计算机科学系，2011年）以及莱亚尔等人（第9届图像分析与识别国际会议论文集，第二卷，ICIAR'12。柏林，海德堡：施普林格出版社；2012年。第432 - 439页）处理此类图像的方法相比，我们得出结论，所提出的方法在利什曼原虫感染的自动标注方面具有更好的性能。
