Abstract:The application of machine learning in wavefront reconstruction has brought great benefits to real-time, non-invasive, deep tissue imaging in biomedical research. However, due to the diversity and heterogeneity of biological tissues, it is difficult to train the dataset with a unified model. In general, the utilization of some unified models will result in the specific sample falling outside the training set, leading to low accuracy of the machine learning model in some real applications. This paper proposes a sensorless wavefront reconstruction method based on transfer learning to overcome the domain shift introduced by the difference between the training set and the target test set. We build a weights-sharing two-stream convolutional neural network (CNN) framework for the prediction of Zernike coefficient, in which a large number of labeled randomly generated samples serve as the source-domain data and the unlabeled specific samples serve as the target-domain data at the same time. By training on massive labeled simulated data with domain adaptation to unlabeled target-domain data, the network shows better performance on the target tissue samples. Experimental results show that the accuracy of the proposed method is 18.5% higher than that of conventional CNN-based method and the peak intensities of the point spread function (PSF) are more than 20% higher with almost the same training time and processing time. The better compensation performance on target sample could have more advantages when handling complex aberrations, especially the aberrations caused by various histological characteristics, such as refractive index inhomogeneity and biological motion in biological tissues.

Sensing morphogenesis of bone cells under microfluidic shear stress by holographic microscopy and automatic aberration compensation with deep learning

Wavefront Reconstruction Based on Deep Transfer Learning for Microscopy.

Machine learning guided rapid focusing with sensor-less aberration corrections.

Adaptive Optics For Structured Illumination Microscopy Based On Deep Learning

Unlabeled flow cellular deformation measurement based on digital holographic microscopy

Simplifying the detection of optical distortions by machine learning

Aberration Correction During Real Time in Vivo Imaging of Bone Marrow with Sensorless Adaptive Optics Confocal Microscope.

Phase-aberration compensation via deep learning in digital holographic microscopy

Deep Learning of Structural Morphology Imaged by Scanning X-ray Diffraction Microscopy

OAH-Net: A Deep Neural Network for Hologram Reconstruction of Off-axis Digital Holographic Microscope

Deep learning to overcome Zernike phase-contrast nanoCT artifacts for automated micro-nano porosity segmentation in bone

Intrinsic Spatial Shift of Local Focus Metric Curves in Digital Inline Holography for Accurate 3D Morphology Measurement of Irregular Micro-Objects

Deep learning-driven adaptive optics for single-molecule localization microscopy

Automatic and accurate compensation for phase aberrations in digital holographic microscopy based on iteratively reweighted least squares fitting

Automatic Segmentation of Osteonal Microstructure in Human Cortical Bone Using Deep Learning: A Proof of Concept

Multiscale methodology for bone remodelling simulation using coupled finite element and neural network computation

Experiment of Structural Geometric Morphology Monitoring for Bridges Using Holographic Visual Sensor

Optical Aberration Correction via Phase Diversity and Deep Learning

Multiscale approach for bone remodeling simulation based on finite element and neural network computation

Deep learning enhanced mobile-phone microscopy

Phase Aberration Separation for Holographic Microscopy by Alternating Direction Sparse Optimization.