Cavendish Laboratory, Department of Physics, University of Cambridge, J.J. Thomson Avenue, Cambridge CB3 0HE, U.K.
ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, Didcot, Oxfordshire OX11 OQX, U.K.
J Chem Inf Model. 2021 Mar 22;61(3):1136-1149. doi: 10.1021/acs.jcim.0c01455. Epub 2021 Mar 8.
Automating the analysis portion of materials characterization by electron microscopy (EM) has the potential to accelerate the process of scientific discovery. To this end, we present a Bayesian deep-learning model for semantic segmentation and localization of particle instances in EM images. These segmentations can subsequently be used to compute quantitative measures such as particle-size distributions, radial- distribution functions, average sizes, and aspect ratios of the particles in an image. Moreover, by making use of the epistemic uncertainty of our model, we obtain uncertainty estimates of its outputs and use these to filter out false-positive predictions and hence produce more accurate quantitative measures. We incorporate our method into the ImageDataExtractor package, as ImageDataExtractor 2.0, which affords a full pipeline to automatically extract particle information for large-scale data-driven materials discovery. Finally, we present and make publicly available the Electron Microscopy Particle Segmentation (EMPS) data set. This is the first human-labeled particle instance segmentation data set, consisting of 465 EM images and their corresponding semantic instance segmentation maps.
通过电子显微镜（EM）自动化材料特性分析部分具有加速科学发现进程的潜力。为此，我们提出了一种贝叶斯深度学习模型，用于对 EM 图像中的粒子实例进行语义分割和定位。这些分割可以随后用于计算定量测量，例如粒子尺寸分布、径向分布函数、平均尺寸和图像中粒子的纵横比。此外，通过利用我们模型的认识不确定性，我们获得了其输出的不确定性估计，并利用这些估计来过滤掉假阳性预测，从而产生更准确的定量测量。我们将我们的方法整合到 ImageDataExtractor 包中，作为 ImageDataExtractor 2.0，它提供了一个完整的流水线，用于自动从大规模数据驱动的材料发现中提取粒子信息。最后，我们提出并公开了 Electron Microscopy Particle Segmentation (EMPS) 数据集。这是第一个人工标记的粒子实例分割数据集，由 465 张 EM 图像及其对应的语义实例分割图组成。
