Deep learning for scanning electron microscopy: Synthetic data for the nanoparticles detection

Deep learning algorithms are one of most rapid developing fields into the modern computation technologies. One of the bottlenecks into the implementation of such advaced algorithms is their requirement for a large amount of manually-labelled data for training. For the general-purpose tasks, such as general purpose image classification/detection the huge images datasets are already labelled and collected. For more subject specific tasks (such as electron microscopy images treatment), no labelled data available. Here I demonstrate that a deep learning network can be successfully trained for nanoparticles detection using semi-synthetic data. The real SEM images were used as a textures for rendered nanoparticles at the surface. Training of RetinaNet architecture using transfer learning can be helpful for the large-scale particle distribution analysis. Beyond such applications, the presented approach might be applicable to other tasks, such as image segmentation.

Cytotoxicity control of silicon nanoparticles by biopolymer coating and ultrasound irradiation for cancer theranostic applications

Silicon nanoparticles (SiNPs) prepared by mechanical grinding of luminescent porous silicon were coated with a biopolymer (dextran) and investigated as a potential theranostic agent for bioimaging and sonodynamic therapy. Transmission electron microscopy, photoluminescence and Raman scattering measurements of dextran-coated SiNPs gave evidence of their enhanced stability in water. In vitro experiments confirmed the lower cytotoxicity of the dextran-coated NPs in comparison with uncoated ones, especially for high concentrations of about 2 mg ml^-1. Efficient uptake of the NPs by cancer cells was found using bioimaging in the optical transmittance and photoluminescence modes. Treatment of the cells with uptaken SiNPs by therapeutic ultrasound for 5-20 min resulted in a strong decrease in the number of living cells, while the total number of cells remained nearly unchanged. The obtained data indicate a 'mild' effect of the combined action of ultrasonic irradiation and SiNPs on cancer cells. The observed results reveal new opportunities for controlling the photoluminescent and sonosensitizing properties of silicon-based NPs for applications in the diagnostics and mild therapy of cancer.