Optical endomicroscope via fiber bundles has been widely used to provide cellular-level visualization for clinical applications. However, improving the quality and spatial resolution of the under-sampling images obtained by fiber bundle with the inter-core spacing remains a challenging problem. In this paper, we address the problem of deconvolution and restoration in fiber-bundle-based imaging. We propose a fast iterative shrinkage-thresholding algorithm (FISTA) solve this problem, and a high-resolution (HR) image without honeycomb patterns is restored from a low-resolution (LR) fiber bundle image. The feasibility of this mothed is verified by experimental results, which shows a promising and wide applications for fiber bundle imaging.
Fiber bundle have been widely used in biomedical endoscopy areas because of its flexibility and high spatial resolution. However, due to the irregular layout of the core and surrounding cladding, the captured fiber bundle images are affected by repeated honeycomb-like patterns. In this paper, we investigate the conditional generative adversarial network to reconstruct the optical fiber endoscope image. In order to improve the operation efficiency and reduce the burden of generative network, we use traditional method to obtain the fiber location map, as an additional condition to generate the network together with the input image. By injecting the position information of fibers, the generative network will pay more attention to the fiber regions and the surrounding structures, and the discriminative network will be able to assess the local consistency of the restored regions. In order to obtain more extensive contextual information, we apply multi-scale loss on the decoder side of the autoencoder. Each of these losses compares the difference between the output of the convolutional layer and the ground truth value that has been downscaled accordingly. Aside from the multi-scale losses, which are based on a pixel-by-pixel operation, we also add a perceptual loss that measures the global discrepancy between the features of the autoencoder’s output and those of the corresponding ground-truth clean image. Our experiments show the effectiveness of our approach which can effectively remove the honeycomb-like patterns and retain the original image features.
High-resolution optical microscope is a crucial imaging equipment in the field of Biomedical Sciences, drug analysis, field investigation and so on. However, the existing commercial microscopes are bulky, cumbersome and expensive, which are only operated by professionals in specific places such as laboratories and have a high threshold for use. In this study, we designed a miniature (20cmx20cmx16cm) lightweight (1.2 kg) low cost (700 USD) inverted optical microscope, the field of view is 0.48mm x 0.38mm, and the resolution can reach 2.19um, we seal the whole imaging system in a very small space, isolated dust and vapor interference from the outside to optical system, to ensure the quality of image and the stability of work. The internal objective lens can be finely focused on the z-axis electronically, ensuring that the microscope accurately focus samples of different thickness. At the same time, low price and stable performance allow small biological laboratories or field teams to use in a variety of harsh environments. In addition, we demonstrated the application of the microscope in different fields, such as imaging living cells in incubators and observing biological samples in the field. Therefore, this portable and economic microscope provides the basic functions of a bulky and expensive microscope at a low cost.
Diffuse speckle contrast analysis (DSCA) measures blood flow in deep tissues by the sensitivity of speckle contrast signals to the displacement of red blood cells (RBCs). Currently, the most common model for describing the displacement of RBCs is a Brownian diffusion-like process. In fact, RBCs undergo shear-induced displacement in blood flow. In this paper, the reduction in speckle contrast due to shear-induced motions is studied by theory and Monte Carlo simulations. We provide the solution for the speckle contrast function in a semi-infinite geometry, and establish the quantitative relationship between speckle contrast and absolute blood flow in a realistic vascular network. Based on this relationship, we can determine the relative contributions of diffusive RBCs motions on the speckle contrast.
A practical technique, known as Confocal laser endoscopy with the ability of super high magnification and sensitivity, uses a pinhole to reduce the stray light from the samples before and after the focal plane, leading to greatly improve the signal-to-noise ratio (SNR) and axial resolution. The algorithm and the computational efficiency of the image acquisition system as one of the key modules of confocal laser endoscopy significantly influence the imaging quality and timeliness of endoscopy. Therefore, it is necessary to design an image acquisition system with good image quality, fast image refreshing and user-friendly interface to improve the use effect of confocal laser endoscopy and the working efficiency for the doctors. In this paper, a high-frame-rate data acquisition system is designed for the scanning and imaging process of confocal laser endoscopy. By using hybrid programming between C# and C++, the acquisition system able to effectively utilize the computational efficiency of C++ language and the flexible graphical interface support of C# language. The WPF framework is used to realize image display and translation scaling, which can help operators to observe samples. In this paper, we have established a test system and tested the related performance indicators. The experimental results show that this image acquisition system of confocal laser endoscopy can achieve a 512×512 pixel imaging with the speed up to 15 frames/s, and correct each frame of the image in real time.
Previously, we presented an on-axis linear-response linear-motion optical scanner. While the linear design is highly desired for engineering consideration, it was still lacking the scanning speed required for imaging applications. We here present a customized profile lens (CPL), tailored for high speed performance while maintaining the advantages of a linear response on-axis optical scanner. The device was built and tested experimentally on an optical bench. The test results demonstrate precise linear response and fast scanning speed, and revealed video frame rate scanning ability. The implementation of the CPLs in laser scanning systems is promising in improving the current 3D laser scanning microscopy systems by reducing the size, error, and complexity of the system, as well as other systems unitizing high speed laser scanning technique.