A method to reduce coherent noise in digital holographic phase contrast microscopy is proposed. By slightly shifting the specimen, a series of digital holograms with different coherent noise patterns is recorded. Each hologram is reconstructed individually, while the different phase tilts of the reconstructed complex amplitudes due to the specimen shifts are corrected in the hologram plane by using numerical parametric lens method. Afterward, the lateral displacements of the phase maps from different holograms are compensated in the image plane by using digital image registration method. Thus, all phase images have same distribution, but uncorrelated coherent noise patterns. By a proper averaging procedure, the coherent noise of phase contrast image is reduced significantly. The experimental results are given to confirm the proposed method.

Digital holography is a technique that provides a non-invasive, label-free, quantitative, and high-resolution imaging employable in biological and science of matter fields, but not only. In the last decade, digital holography (DH) has undergone very significant signs of progress that made it one of the most powerful metrology tools. However, one of the most important issues to be afforded and solved for obtaining quantitative phase information about the analyzed specimen is related to phase aberrations. Sources of aberrations can be diverse, and several strategies have been developed and tested to make DH a reliable optical system with submicron resolution. This paper reviews the most effective and robust methods to remove or compensate phase aberrations in retrieved quantitative phase imaging by DH. Different strategies are presented and discussed in detail on how to remove or compensate for such disturbing aberrations. Among the various methods improvements in the optical setups are considered the numerical algorithms, the hybrid methods, and the very recent Artificial Intelligence (AI) approaches to compensate for all aberrations which affect the setups to improve the imaging quality and the accuracy of the reconstruction images’ procedures.

We propose a method to compensate the aberrations of digital holographic microscopy based on a multiple regression convolutional neural network. The experimental verification is carried out by measuring the morphological change of living cells under micro-Fluidic Shear Stress.

An aperture synthesis approach of digital holography for microscopy imaging at long working distance is proposed. Firstly, for an oblique object, a series of Fresnel off-axis holograms are recorded with different tilted plane wave illuminations without using lens for pre-magnification. Then the complex amplitudes are reconstructed and magnified from these holograms by the double-step Fresnel reconstruction method respectively. Finally, the synthesized image of the resolution enhanced and the speckle suppressed is obtained by incoherent superposition of these reconstructed complex amplitudes. The important advantage of the proposed approach is that the working distance of the system isn't constrained and the reconstructed image doesn't subject to lens aberrations. The experimental results with a die and an USAF-1951 resolution test target are shown and demonstrated that the resolution of both intensity and phase image can be effectively enhanced with simple set-up and procedure. The proposed approach can improve the capabilities of digital holography in three-dimensional in-situ microscopy at long working distance.

About 75% of epithelial ovarian cancer (EOC) patients suffer from relapsing and develop drug resistance after primary chemotherapy. The commonly used clinical examinations and biological tumor tissue models for chemotherapeutic sensitivity are time-consuming and expensive. Research studies showed that the cell morphology-based method is promising to be a new route for chemotherapeutic sensitivity evaluation. Here, we offer how the drug resistance of EOC cells can be assessed through a label-free and high-throughput microfluidic flow cytometer equipped with a digital holographic microscope reinforced by machine learning. It is the first time that such type of assessment is performed to the best of our knowledge. Several morphologic and texture features at a single-cell level have been extracted from the quantitative phase images. In addition, we compared four common machine learning algorithms, including naive Bayes, decision tree, K-nearest neighbors, support vector machine (SVM), and fully connected network. The result shows that the SVM classifier achieves the optimal performance with an accuracy of 92.2% and an area under the curve of 0.96. This study demonstrates that the proposed method achieves high-accuracy, high-throughput, and label-free assessment of the drug resistance of EOC cells. Furthermore, it reflects strong potentialities to develop data-driven individualized chemotherapy treatments in the future.

In optical tomography, it is challenging to obtain high-quality results for complex-structured objects which induce multiple scattering. Nonlinear reconstruction methods outperform linear ones in these situations. A promising nonlinear method is the approach based on beam propagation method, but its accuracy may decrease for complicated structures. In this paper, we describe a novel tomographic reconstruction method using multi-slice wave propagation method (WPM) as the forward model, which simulates the scattering process more precisely but has not been introduced in tomographic reconstruction before. The computational model of WPM is presented. To tackle the computational complexity, we propose an efficient scheme to compute the transmitted field and its derivative. We then use an iterative optimization method to recover the quantitative refraction index distribution. We also discuss the influences of the parameters in the method and how to determine their values. The experimental results demonstrate that this method can address multiple scattering problems and provide high accuracy for complex-structured objects.