Gene detection method and device based on deep learning and fluorescence spectrum

Abstract

The invention relates to a gene detection method and device based on deep learning and fluorescence spectra. The method comprises the steps: acquiring fluorescence spectrum images of different markersof a plurality of genes, randomly selecting M fluorescence spectrum images of different wavebands for each gene, fusing the fluorescence spectrum images into a mixed fluorescence spectrum image, enhancing the fluorescence spectrum images; filtering background noise of the mixed spectrum image according to Gaussian peak hypothesis and a local adaptive polynomial fitting algorithm, and then performing feature extraction on a second mixed spectrum image according to a maximum value and minimum value adaptive algorithm to obtain peak signal features of the fluorescence spectrum images; and training a convolutional neural network model according to the data, and acquiring a gene detection result by using the convolutional neural network model. According to the method, traditional filtering andimage processing are combined, and the feature extraction is carried out on the fluorescence spectrum images; and the fluorescence spectra of different markers in the sample improve the robustness, generalization ability and accuracy of the convolutional neural network model.

Description

technical field [0001] The invention relates to the fields of biological information and deep learning, in particular to a gene detection method and device based on deep learning and fluorescence spectroscopy. Background technique [0002] Fluorescence-based detection methods are extremely important analytical methods in analytical chemistry, including fluorescence excitation / emission spectroscopy, phase-resolved fluorescence analysis, time-resolved fluorescence analysis, fluorescent immunolabeling analysis, three-dimensional fluorescence analysis, etc. Fluorescence analysis is widely used in environmental analysis, medical analysis, and biological imaging due to its high sensitivity, good selectivity, wide linear working range, and the ability to easily adapt to analysis requirements through chemical means such as synthesis and modification. , genetic testing and other fields. [0003] The accuracy and speed of fluorescence-based gene detection depend on the specificity of...

AI Technical Summary

Problems solved by technology

[0003] The accuracy and speed of fluorescence-based gene detection depend on the specificity of the combination of gene markers (nucleotide fragments) and targeted gene fragments. In the case of some unknown or newly discovered gene fragments, it is impossible to obtain specific On the other hand, due to the latent failure or improper operation of the detection equipment, the sample is polluted, the characteristic wavelength signal is weak, or the background noise is large, or other problems, which make the detection result difficult. Determined or large deviation

Images