In this study, we describe a simple method to produce signals which can reveal the cross-sectional information of samples in an optical coherence tomography (OCT) system. Instead of using the spectrometer and the Fourier transformation calculation in the conventional spectrum domain (SD) OCT system, we use a Mach-Zehnder interferometer structure of the spatial heterodyne spectrometer. In a spatial heterodyne spectrometer, because each position on the photodetector array could be mapped to a specific optical path difference, the spectral density distribution could be retrieved with Fourier transformation. And in an SD-OCT system, cross-section signals are obtained by conducting Fourier transformation to the spectrum signals. Therefore, in our OCT system, the spatial signals captured by the photodetector array is related to the cross-sectional signals obtained in an SD-OCT system. The theoretical study and the numerical simulation demonstrate that by applying our method in an OCT system, the heterodyne spectrometer structure could generate a symmetrical pattern composed of fringes with high spatial frequency. Then the photodetector array captures the pattern to generate a spatial signal. The spatial ordinate of this signal is linearly related to the optical depth in sample, while the amplitude of the signal intensity variation is linearly related to the intensity of coherent backscattered light in the sample. The imaging depth is theoretically unlimited. Also, because of the high spatial frequency of the signal, we further adjust the inclination angle in the heterodyne spectrometer structure to visualize the signal.
Spectral analysis is an important method for noninvasive blood glucose measurement. Presently, Fourier-transform spectroscopy is a well-established technique that provides highly resolved spectral measurements in the infrared, visible and ultraviolet ranges. In this study, we proposed a novel method for obtaining linear spectra based on regular Spatial Heterodyne Spectrometers. In particular, we wanted to use a fluorescent dye-coated screen and a Fourier lens to directly obtain uniform K-space spectra. In the system, the up-conversion luminescent material on the screen is hoped to absorb coherent incident light and emit light of a specific wavelength that maintains the coherence. According to our calculation, the photodetector array receives the Fourier image pattern on the screen and can directly obtain the spectrum of the measured substance, therefore the scientists can directly observe the spectrum of the test sample. Furthermore, we replace the fluorescent dye-coated screen by an infrared laser detector card, which is commonly used in laboratories, to primary verify the feasibility of the method. Up-conversion luminescent materials that are widely used in the fields of analytical chemistry, biomedicine, and life sciences, have very good application prospects in biological imaging, photodynamic therapy, solar cells, flexible fluorescent films and sensing.
We present an automatic classification algorithm for retinal optical coherence tomography (OCT) images based on convolution neural network (CNN). This algorithm inherently contains feature extraction and classification, thus avoiding the design feature extractor manually. Firstly, we processed the OCT images to focus on and determine the pathological area of the retinal OCT images, and to speed up the training of the network. Then we input the original images to crop them, which can effectively prevent the noise introduced in the processes of image processing and changing the pixels in the original image. Secondly, we augmented the OCT images in the source data set to obtain sufficient images, and to alleviate the impact of a relatively small number of target classification images on the model accuracy and generalization ability. Our method was introduced the random translation in image cropping and horizontal flipped to augment the OCT images. Then we applied two methods to build two data sets used to train the network, and we divided each of the data sets into a training set and a validation set. Next, we designed an efficient classification network and trained it with the two training sets respectively, to acquire the two models that can classify OCT images. The results indicate that the network trained by the augmented data can classify images more effectively. In our classification algorithm, the accuracy, the sensitivity and the specificity are 93.43%, 91.38%, and 95.88%, respectively.
Diabetes has been a serious problem that poses threat to people's health all around the world. It is still a challenge for us to detect blood glucose concentration continuously and non-invasively. In this research, we developed a free-space spectrum domain optical coherence tomography (SD-OCT) system for non-invasive blood glucose detection which possessed advantages of easy construction, analyzation and control. In this system, a laser with center wavelength of 980nm was applied because of its low absorption in both glucose and water, which was suitable for OCT imaging. However, the laser with wavelength of 980nm was not used in the OCT with optic fiber type which was commercially designed for wavelengths of 830nm, 1310nm and 1550nm. By applying a dispersing prism, we could obtain higher resolution spectrum to acquire better OCT images and more accurate glucose concentration. The tomography function of this free-space SD-OCT system was proved to work by scanning onion sample. Pristella maxillaris is a kind of fish with transparent body structure and suitable size, thus we consider it to be an ideal animal for blood glucose measurement by optical methods. We cultivated pristella maxillaris, an ideal fish for this experiment, in glucose solutions with five different concentrations as samples to study glucose monitoring. The OCT signals of the five groups correlated respectively to the glucose concentrations. Therefore, our method provided the potential for measuring blood glucose concentration non-invasively.