Electron Microscopy Analysis Station, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan.
Materials Data Platform Center, 1-1 Namiki Tsukuba Ibaraki 305-0044, Japan.
Microscopy (Oxf). 2022 Jun 6;71(3):161-168. doi: 10.1093/jmicro/dfac008.
It is difficult to discriminate the amorphous state using a transmission electron microscope (TEM). We discriminated different amorphous states on TEM images using persistent homology, which is a mathematical analysis technique that employs the homology concept and focuses on 'holes'. The structural models of the different amorphous states, that is, amorphous and liquid states, were created using classical molecular dynamic simulation. TEM images in several defocus conditions were simulated by the multi-slice method using the created amorphous and liquid states, and their persistent diagrams were calculated. Finally, logistic regression and support vector classification machine learning algorithms were applied for discrimination. Consequently, we found that the amorphous and liquid phases can be discriminated by more than 85%. Because the contrast of TEM images depends on sample thickness, focus, lens aberration, etc., radial distribution function cannot be classified; however, the persistent homology can discriminate different amorphous states in a wide focus range.
使用透射电子显微镜（TEM）很难区分无定形状态。我们使用持久同调（一种数学分析技术，采用同源概念，专注于“孔”）来区分 TEM 图像上的不同无定形状态。不同无定形状态（即无定形和液体状态）的结构模型是使用经典分子动力学模拟创建的。使用创建的无定形和液体状态，通过多片方法模拟了几个离焦条件下的 TEM 图像，并计算了它们的持久图。最后，应用逻辑回归和支持向量分类机器学习算法进行判别。结果表明，无定形相与液相的判别率超过 85%。由于 TEM 图像的对比度取决于样品厚度、焦点、透镜像差等，因此无法对径向分布函数进行分类；然而，持久同调可以在较宽的焦点范围内区分不同的无定形状态。
