小波变换在量子点编码识别中的应用*☆

背景：为实现更多种类的量子点编码，量子点的荧光发射峰必定出现重叠，这就造成了量子点编码识别的困难。 目的：应用小波变换对重叠峰展开技术，对两种相邻波长量子点的编码进行识别。 方法：在显微镜引导下，以375 nm的紫外光激发单个量子点编码微球的荧光光谱，被光纤光谱仪采后，应用小波变换对数据多维展开，然后经过样条函数处理，再通过小波反变换重构波谱展开的光谱。 结果与结论：为了使处理数据的长度为2n，在原始数据进行了插值运算。经过小波变换处理后，峰位置保持不变，能够提取编码荧光光谱的特征值。通过小波变换，混合微球的荧光光谱被展开，能够识别出两种量子点编码，提高了识别的分辨率。通过提高对相邻光谱的编码的识别，量子点编码的颜色必将增加，从而显著丰富编码的信息量。结果提示用小波变换对光谱进行处理后，可以实现对不同波长的荧光展开，提高了识别效率，为实现肿瘤标志物的高通量检测奠定基础。

Application of Wavelet in Quatum Dot Barcodes Recognition

BACKGROUND: To obtain more quantum dot (QD) barcodes, the overlay peaks of fluorescence occur, leading to difficulties in identifying QD barcodes. OBJECTIVE: To identify QD barcodes of two adjacent wave length using wavelet transform technique. METHODS: Through the microscopy, the spectrum of fluorescence induced by 375 nm light was captured by spectroscopy. The spectral signal was split into multi-scale components by wavelet transform. After transformed by spline function, every component constructed a new spectrum with peaks expanded by inverse wavelet transform. RESULTS AND CONCLUSION: Interpolation operation was performed on original data to control the data length to 2n. Following wavelet transform, peak location remained unchanged, so the eigenvalue of spectrum of coding fluorescence was extracted. The spectrum of fluorescence mixed with microspheres was split, and two QD barcodes were identified. The improved barcodes identification of adjacent spectrum increase color of QD barcodes, thereby enhancing code information volume. Results show that following spectrum was processed by wavelet transform, overlay peaks of fluorescence has be expanded, and enhanced the efficiency of recognition, which lays a foundation for detecting tumor markers.