Systematic mapping: Use of neural networks for analysis in transmission electron microscopy micrographs

Microscopy techniques have been prominently part of advances in fields such as biology, medicine, and the study and development of materials over the last decade. The characteriza- tion of nanoparticles from distinct materials is challenging due to differences in morphology, size and shape. In order to evaluate these properties, tools such as the optical microscope, the atomic force microscope and the transmission electron microscope are essential. However, even an apparently simple measurement such as the size of a particle can be a challenge, this task becomes more difficult if you are working with materials for which particles are far from the ideal shape. In recent years, artificial neural networks (RNA) and especially convolutional neural networks (RNC) have been showing an enormous capacity for complex vision tasks such as detection, segmentation and classification.In this paper, we show the results of a mapping study on the work related to implementations of RNA for the analysis of electron transmission microscope micrographs. In addition to reviewing the techniques and tools most commonly used in RNA implementations. Results indicate a growing interest of the scientific community to propose related solutions, with RNC as the leader technique in micrograph analysis.