Fig. 1: Basic characteristics of AI-nanopore platform.
a Nanopore structure fabricated in silicon nitride (SiN) membrane on a silicon (Si) substrate. A specimen and a buffer are placed in the cis and trans channels. Silver–silver chloride (Ag/AgCl) electrodes are placed on both sides of the substrate. b Ionic current flows when voltage is applied. Ip and td show the peak current and duration of the current of an ionic current waveform, respectively. c An optical photographic image of the nanopore module. Red and blue are cis and trans channels, respectively. The black central part is a silicon chip with nanopores. d, e Scanning electron microscope image of the nanopore entrance and the nanopore with a diameter of 300 nm seen from the trans side. The black region in the center of the image is a hollow SiN thin film. Six times experiments were repeated independently for the observations. f Ionic current–time traces of nanoparticles with 200 and 220 nm diameter. g, h Ionic current–time profiles, which show an overlap of 100 waveforms each, obtained by nanoparticle measurements. The color code corresponds to f. i, and j, Histograms of Ip and td of nanoparticles. The color code corresponds to f. k, Machine learning algorithm used to identify the nanoparticles and cultured viruses. l Confusion matrix of nanoparticles obtained by machine learning. The numbers in the matrix represent the number of waveforms obtained by the measurements. The color bar indicates that the darker the blue, the greater the number of pulses. The number of the color bar indicated the number of waveforms. Source data of f–j and l are provided as a Source Data file.
