Velmurugan Ramraj, Chao Jerry, Ram Sripad, Ward E Sally, Ober Raimund J
Opt Express. 2017 Feb 20;25(4):3394-3410. doi: 10.1364/OE.25.003394.
Multifocal plane microscopy (MUM) can be used to visualize biological samples in three dimensions over large axial depths and provides for the high axial localization accuracy that is needed in applications such as the three-dimensional tracking of single particles and super-resolution microscopy. This report analyzes the performance of intensity-based axial localization approaches as applied to MUM data using Fisher information calculations. In addition, a new non-parametric intensity-based axial location estimation method, Multi-Intensity Lookup Algorithm (MILA), is introduced that, unlike typical intensity-based methods that make use of a single intensity value per data image, utilizes multiple intensity values per data image in determining the axial location of a point source. MILA is shown to be robust against potential bias induced by differences in the sub-pixel location of the imaged point source. The method's effectiveness on experimental data is also evaluated.
多焦平面显微镜（MUM）可用于在较大轴向深度上对生物样本进行三维可视化，并为单粒子三维跟踪和超分辨率显微镜等应用所需的高轴向定位精度提供支持。本报告使用费舍尔信息计算分析了应用于MUM数据的基于强度的轴向定位方法的性能。此外，还介绍了一种新的基于强度的非参数轴向位置估计方法——多强度查找算法（MILA），与典型的基于强度的方法不同，后者在每个数据图像中使用单个强度值，而MILA在确定点源的轴向位置时，在每个数据图像中利用多个强度值。结果表明，MILA对成像点源亚像素位置差异引起的潜在偏差具有鲁棒性。还评估了该方法对实验数据的有效性。
