Journal: RSC Advances

Article Title: Toward sustainable diagnostics for Candida albicans : the role of artificial intelligence in analytical chemistry from data processing to Python-based blueness and redness evaluation metrics

doi: 10.1039/d6ra00286b

Figure Lengend Snippet: Analytical methods for detection of candida albicans

Article Snippet: 2. Bioconjugate concanavalin A to CdTe-MSA QDs with specific labeling of on hyphae and yeast C. albicans cells. , 2. Labeling with QDs-(Con A) conjugated to antibodies with high-resolution labeling of biofilms both in vitro and in vivo . , 2. Conjugation: QDs : Con A ratio 1000 : 1, pH 8.0, 2 h incubation at room temperature with gentle stirring. , 2. UV-vis Absorption spectroscopy: spectrophotometer (thermo scientific)..

Techniques: Labeling, Fluorescence, Spectroscopy, Cell Culture, Suspension, Circular Dichroism, Confocal Microscopy, Flow Cytometry, In Vitro, In Vivo, Conjugation Assay, Incubation, Gentle, Spectrophotometry, Concentration Assay, Isolation, Raman Spectroscopy, Metabolic Labelling, Bacteria, Diagnostic Assay, Colorimetric Assay, Binding Assay, Sample Prep, Software, Biomarker Discovery, Infection, In Situ Hybridization, Sequencing, Microscopy, Hybridization, Transmission Assay, Enzyme-linked Immunosorbent Assay, Centrifugation, Nuclear Magnetic Resonance, Structural Proteomics, Derivative Assay, Selection, Electrophoresis, Red Blood Cell Lysis, Quantitation Assay, Mass Spectrometry, Gas Chromatography, Disruption, Produced, Impedance Spectroscopy, Activation Assay, Blocking Assay, Activity Assay, Construct, Förster Resonance Energy Transfer, Real-time Polymerase Chain Reaction, Saline, Clinical Proteomics, Membrane, Functional Assay

Journal: Science Advances

Article Title: Electroluminescent perovskite QD–based neural networks for energy-efficient and accelerate multitasking learning

doi: 10.1126/sciadv.ady8518

Figure Lengend Snippet: ( A ) Conceptual representations of facial images with ethnic and age diversity. ( B ) Robot performing MT learning based on various face images for tasks 1 to 3. ( C ) Dual-output synaptic device array generating PSC and PSEL for MT learning. ( D ) Junction constituents of the dual-output Cs 1− x FA x PbBr 3 QD–based synaptic device cell with a cross-sectional HR-TEM image. ( E ) STEM-EDX analysis with elemental mapping of indium (In), zinc (Zn), bromine (Br), carbon (C), vanadium (V), Al, and platinum (Pt) (from left to right) of the Cs 1− x FA x PbBr 3 QD–based synaptic device. ( F ) (i) HR-TEM image of the Cs 1− x FA x PbBr 3 ( x = 0.047) QDs. (ii to vi) Optical images of the electroluminescent (green color) Cs 1− x FA x PbBr 3 (0.00 ≤ x ≤ 0.15) QD–based synaptic array (64 cells). The active area is 100 μm by 100 μm. ( G ) CIE 1931 coordinates of the generated EL emission, exhibiting higher X and Y coordinates as the x increases.

Article Snippet: Time-resolved EL (TR-PSEL, fig. S20C) and Tr.-PSEL (fig. S24) measurements were accomplished by using a function generator (Keysight; 81150A) to generate a square-wave signal, a photomultiplier (Hamamatsu; H10721-01) to receive lights generated by the Cs 1− x FA x PbBr QD–based synaptic device, and an oscilloscope (Keysight; DSOX1204G) to receive the square-wave signal and light-emitting response signal of the devices.

Techniques: Generated

Journal: Science Advances

Article Title: Electroluminescent perovskite QD–based neural networks for energy-efficient and accelerate multitasking learning

doi: 10.1126/sciadv.ady8518

Figure Lengend Snippet: ( A ) PSEL spectra as a function of V P and wavelength at x = 0.15. A.U., arbitrary units. ( B ) PSEL spectra as a function of P W and wavelength at x = 0.15. ( C ) PSEL spectra as a function of PN and wavelength at x = 0.15. ( D ) I - V curve at x = 0.11. Responses of ( E ) PSC and ( F ) PSEL triggered by a pair of presynaptic electrical spikes at Δ t = 0.1 ms. ( G ) PSEL responses (middle) and statistical heatmaps for I ON and I OFF (top) of 64 cells of the Cs 1− x FA x PbBr 3 QD–based synaptic array (bottom; optical image of the 8-by-8 synaptic array). ( H ) PPF characteristics as a function of Δ t for PSC (blue) and PSEL (red) responses. The solid lines are the fitted curves. ( I ) Retention of PSC and PSEL as functions of t and PN. ( J ) Learning-memory-forgetting process driven by temporal evolution of PSC and PSEL responses under repeated pulse stimulation.

Article Snippet: Time-resolved EL (TR-PSEL, fig. S20C) and Tr.-PSEL (fig. S24) measurements were accomplished by using a function generator (Keysight; 81150A) to generate a square-wave signal, a photomultiplier (Hamamatsu; H10721-01) to receive lights generated by the Cs 1− x FA x PbBr QD–based synaptic device, and an oscilloscope (Keysight; DSOX1204G) to receive the square-wave signal and light-emitting response signal of the devices.

Techniques:

Journal: Science Advances

Article Title: Electroluminescent perovskite QD–based neural networks for energy-efficient and accelerate multitasking learning

doi: 10.1126/sciadv.ady8518

Figure Lengend Snippet: ( A ) LTP and LTD functions of PSC responses as a function of x (0.00 ≤ x ≤ 0.15). ( B ) DR as functions of x and PN. ( C ) LTP and LTD functions of PSC responses as a function of PN (# = 35, 80, 130, 530, and 1000) at x = 0.11. ( D ) Evolution of heatmap visualization of the “T” shape in a 3-by-3 array as a function of PN during 530 LTP (top) and 530 LTD (bottom) pulses. ( E ) Configuration of readout and programming voltage inputs for LTP and LTD functions. ( F ) LTP and LTD functions of PSEL responses as a function of x (0.00 ≤ x ≤ 0.15). ( G ) DR (driven by PSEL) as functions of x and PN. ( H ) LTP and LTD functions of PSEL responses as a function of PN (# = 35, 80, 130, 530, and 1000) at x = 0.11. ( I ) Optical images of the Cs 1− x FA x PbBr 3 QD–based 3-by-3 synaptic array illuminating the “T” shape. The schematic of the illuminated cell node is depicted in the bottom of (I). ( J ) Optical images of the illuminated “T” shape in the 3-by-3 array as a function of PN during 530 LTP (top) and 530 LTD (bottom) pulses.

Article Snippet: Time-resolved EL (TR-PSEL, fig. S20C) and Tr.-PSEL (fig. S24) measurements were accomplished by using a function generator (Keysight; 81150A) to generate a square-wave signal, a photomultiplier (Hamamatsu; H10721-01) to receive lights generated by the Cs 1− x FA x PbBr QD–based synaptic device, and an oscilloscope (Keysight; DSOX1204G) to receive the square-wave signal and light-emitting response signal of the devices.

Techniques:

Journal: Science Advances

Article Title: Electroluminescent perovskite QD–based neural networks for energy-efficient and accelerate multitasking learning

doi: 10.1126/sciadv.ady8518

Figure Lengend Snippet: ( A ) Cognitive-inspired MT processing with PSC and PSEL outputs for tasks 1 to 3. All matrix elements are assumed to correspond to the conductance value of the node, and the experimental electrical and optical weights of the dual-output Cs 1− x FA x PbBr 3 QD–based synaptic 8-by-8 array are used as kernels. ( B and D ) MT-CNN model configuration for tasks 1 and 2. It consists of an input layer, four PSEL/PSC convolution layers, max pooling layers, and FC layers (FC1** and FC2**). ( C and E ) MT-ResNet-15 model architecture for tasks 2 and 3. It consists of an input layer, four RB layers, three UB layers, max pooling layers, and an FC2 layer. The ResNet-7–based decoder was used to reconstruct the input image. The total loss for each configuration is determined by considering the relative ratio of regression (reconstruction) loss and classification (regression) loss, which can be used to update the weights of both the shared kernel and task-specific layers by backpropagation.

Article Snippet: Time-resolved EL (TR-PSEL, fig. S20C) and Tr.-PSEL (fig. S24) measurements were accomplished by using a function generator (Keysight; 81150A) to generate a square-wave signal, a photomultiplier (Hamamatsu; H10721-01) to receive lights generated by the Cs 1− x FA x PbBr QD–based synaptic device, and an oscilloscope (Keysight; DSOX1204G) to receive the square-wave signal and light-emitting response signal of the devices.

Techniques: