DeepCell Kiosk: scaling deep learning-enabled cellular image analysis with Kubernetes.

Deep learning is transforming the analysis of biological images, but applying these models to large datasets remains challenging. Here we describe the DeepCell Kiosk, cloud-native software that dynamically scales deep learning workflows to accommodate large imaging datasets. To demonstrate the scalability and affordability of this software, we identified cell nuclei in 106 1-megapixel images in ~5.5 h for ~US$250, with a cost below US$100 achievable depending on cluster configuration. The DeepCell Kiosk can be downloaded at https://github.com/vanvalenlab/kiosk-console ; a persistent deployment is available at https://deepcell.org/ .

Quantum Ensembles of Silicon Nanoparticles: Discrimination of Static and Dynamic Photoluminescence Quenching Processes.

Porous silicon photoluminescence is characterized by a broad emission band that displays unusually long (tens to hundreds of micro-seconds), wavelength-dependent emissive lifetimes. The photoluminescence is associated with quantum confinement of excitons in silicon nanocrystallites contained within the porous matrix, and the broad emission spectrum derives from the wide distribution of nanocrystallite sizes in the material. The longer emissive lifetimes in the ensemble of quantum-confined emitters correspond to the larger nanocrystallites, with their longer wavelengths of emission. The quenching of this photoluminescence by aromatic, redox-active molecules aminochrome (AMC), dopamine, adrenochrome, sodium anthraquinone-2-sulfonate, benzyl viologen dichloride, methyl viologen dichloride hydrate, and ethyl viologen dibromide is studied, and dynamic and static quenching mechanisms are distinguished by the emission lifetime analysis. Because of the dependence of the emission lifetime on emission wavelength from the silicon nanocrystallite ensemble, a pronounced blue shift is observed in the steady-state emission spectrum upon exposure to dynamic-type quenchers. Conversely, static-type quenching systems show uniform quenching across all emission wavelengths. Thus, the difference between static and dynamic quenching mechanisms is readily distinguished by ratiometric photoluminescence spectroscopy. The application of this concept to imaging of AMC, the oxidized form of the neurotransmitter dopamine that is of interest for its role in neurodegenerative diseases, is demonstrated. It is found that static electron acceptors result in no ratiometric contrast, while AMC shows a strong contrast, allowing ready visualization in a 2-D imaging experiment.