Search

Three-dimensional (3D) observation of a biological sample using serial-section electron microscopy is widely used. However, organelle segmentation requires a significant amount of manual time. Therefore, several studies have been conducted to improve organelle segmentation’s efficiency. One such promising method is 3D deep learning (DL), which is highly accurate. However, the creation of training data for 3D DL still requires manual time and effort. In this study, we developed a highly efficient integrated image segmentation tool that includes stepwise DL with manual correction. The tool has four functions: efficient tracers for annotation, model training/inference for organelle segmentation using a lightweight convolutional neural network, efficient proofreading and model refinement. We applied this tool to increase the training data step by step (stepwise annotation method) to segment the mitochondria in the cells of the cerebral cortex. We found that the stepwise annotation method reduced the manual operation time by one-third compared with the fully manual method, where all the training data were created manually. Moreover, we demonstrated that the F1 score, the metric of segmentation accuracy, was 0.9 by training the 3D DL model with these training data. The stepwise annotation method using this tool and the 3D DL model improved the segmentation efficiency of various organelles.

In this study, we investigated use of local surface plasmon resonance (LSPR) of metal nanoparticles (NPs) as a correlative light and electron microscopy (CLEM) tag for biological samples. Gold NPs in ultra-thin sections for TEM revealed that LSPR could be observed by optical microscopy at sizes of 20 nm or larger. Gold NPs at sizes less than 20 nm could be observed using the gold enhancement method. Therefore, this CLEM tag could be applied to immunoelectron microscopy using this gold enhancement method.

Segmentation of three-dimensional (3D) electron microscopy (EM) image stacks is an arduous and tedious task. Deep convolutional neural networks (CNNs) work well to automate the segmentation; however, they require a large training dataset, which is a major impediment. In order to solve this issue, especially for sparse segmentation, we used a CNN with a minimal training dataset. We segmented a Cerebellar Purkinje cell from an image stack of a mouse Cerebellum cortex in less than two working days, which is much shorter than that of the conventional method. We concluded that we can reduce the total labor time for the sparse segmentation by reducing the training dataset.

Uranyl acetate (UAc) has been generally used not only as a superb staining reagent for ultrathin sections of plastic-embedded biological materials, but also as high-contrast negative stains for biological macromolecules such as particles of protein or virus. However, the use and purchase of radioactive UAc have been restricted. In this study, we determine the performance of ytterbium triacetate, lutetium triacetate, samarium triacetate and gadolinium triacetate as new staining reagents for biological electron microscopy. We observed chemically fixed spinach (Spinacia oleracea) leaves stained with these reagents. Ultrathin sections were stained with these reagents. Some of them were counterstained with lead citrate. The transmission electron microscopy contrast of spinach organelles was evaluated in sections exposed to the conventional stain and new stains. We show acetate salts of samarium, gadolinium, ytterbium and lutetium could be excellent substitutes for UAc for thin section staining and for negative staining. In addition, each reagent showed appreciable negative-staining effects.