The concentration of pepsinogen II in human serum is an important index for the detection of atrophic gastritis. In recent years, the serum pepsinogen tests are proved to be a screening way for gastric carcinoma. A portable microscope imaging system combined with lateral flow strips labeled by fluorescent microsphere for quantitative detection of pepsinogen II was developed, which use immunochromatographic assays. The imaging system used 365-nm ultraviolet LED as the excitation light source, and captured the test strip images through portable microscope. A modified two-dimensional maximum entropy segmentation algorithm is applied to analyze the result. Human serum of 8 different PGII concentrations is used as samples. 10 repeated experiments are carried out on each serum of PGII concentration, and the CV values for the data of each PGII concentration are less than 5%. The fitting curve is drawn according to the experimental data, and the fitting degree is 99.67%. Afforded high linearity within PGII levels of 5.62–240 ug/mL (R2= 0.9919),the limit of detection is 4.72ug/ml. It demonstrated that the device could realize rapid, stable detection for PGII concentration in human serum.
In order to further eliminate aberration and improve resolution, the paper employs parabolic mirror as the collimating mirror and the focusing mirror to design “Z” configuration and “U” configuration optical structure of parabolic spectrometer with the F number 2.5 and the spectral range varying from 250 nm to 850 nm. We conduct experiments on ZEMAX to simulate and optimize the initial parameters of two structures with the root-mean-square (RMS) radius of spots along Y axis as the optimization goal. Through analyzing the spot diagram and the root-mean-square (RMS) of Y axis, we can see that the “U” configuration spectrometers can achieve much better spectral resolution than the “Z” configuration.
Light emitting diode (LED) is widely employed in industrial applications and scientific researches. With a spectrometer, the chromaticity of LED can be measured. However, chromaticity shift will occur due to the broadening effects of the spectrometer. In this paper, an approach is put forward to bandwidth correction for LED chromaticity based on Levenberg-Marquardt algorithm. We compare chromaticity of simulated LED spectra by using the proposed method and differential operator method to bandwidth correction. The experimental results show that the proposed approach achieves an excellent performance in bandwidth correction which proves the effectiveness of the approach. The method has also been tested on true blue LED spectra.
Because of the substrate back reflectance phenomena, the reflectance of optical thin film stack on a transparent substrate is totally different from that of on an opaque substrate. In this paper, a method for the measurement of low reflectance optical film thickness that has substrate back reflectance is proposed for the first time. Through the analysis of the actual substrate back reflectance, a compensation model is introduced to reduce the influence of substrate back reflectance. The experimental results show good fitting precision and proves that this model can be used directly for the measurement of the optical film thickness with substrate back reflectance, and no extra process is needed.
The charge-coupled device (CCD) array spectrometers are increasingly being used in wide variety of scientific researches and industrial applications. However, all CCD detectors suffer some amount of non-linear behavior on response to light, and the accuracy of the CCD array spectrometer measurement will be influenced from the non-linear behavior, the detectable error is presented. Therefore, the non-linearity correction method is important to obtain the accurate results of spectrometers based on the CCD. Here, we proposed a convenient experiment and calculation method to solve the problem of non-linearity. With the combined values of all the pixels across the detector, a 7th order polynomial is fitted in the relation between the normalized counts per second and counts, and the correction coefficients were generated by this polynomial for the pixels. The method to apply the correction is dividing the original response by the calculated correction coefficients for all the pixels. Finally, the CCD detector response is linear to >99.5% after correcting for the non-linearity of spectrometers, experimental results show that the proposed method is reasonable and efficient.