Scanning electron microscopy image representativeness: morphological data on nanoparticles

A sample of a nanomaterial contains a distribution of nanoparticles of various shapes and/or sizes. A scanning electron microscopy image of such a sample often captures only a fragment of the morphological variety present in the sample. In order to quantitatively analyse the sample using scanning electron microscope digital images, and, in particular, to derive numerical representations of the sample morphology, image content has to be assessed. In this work, we present a framework for extracting morphological information contained in scanning electron microscopy images using computer vision algorithms, and for converting them into numerical particle descriptors. We explore the concept of image representativeness and provide a set of protocols for selecting optimal scanning electron microscopy images as well as determining the smallest representative image set for each of the morphological features. We demonstrate the practical aspects of our methodology by investigating tricalcium phosphate, Ca-3(PO4)(2), and calcium hydroxyphosphate, Ca-5(PO4)(3)(OH), both naturally occurring minerals with a wide range of biomedical applications. Lay description A typical sample of a nanomaterial contains a distribution of nanoparticles of various shapes and/or sizes. A single scanning electron microscopy (SEM) image of such a sample often captures only a fragment of the sample, and therefore only a fragment of the morphological variety present in the sample. In order to obtain more complete information about the true sample morphology, one needs to asses the content of a series of SEM images. In our article, we present a framework for extracting morphological information contained in SEM images using computer vision algorithms, and for converting them into numerical particle descriptors representing the particle morphology. We then explore the concept of image representativeness and provide a set of protocols for selecting optimal SEM images as well as determining the smallest representative image set for each of the morphological features. We demonstrate the practical aspects of our methodology by investigating SEM images of a tricalcium phosphate sample, a naturally occurring mineral with a wide range of biomedical applications.