Deep-learning-based cross-modality translation from Stokes image to bright-field contrast.

Significance: Mueller matrix (MM) microscopy has proven to be a powerful tool for probing microstructural characteristics of biological samples down to subwavelength scale. However, in clinical practice, doctors usually rely on bright-field microscopy images of stained tissue slides to identify characteristic features of specific diseases and make accurate diagnosis. Cross-modality translation based on polarization imaging helps to improve the efficiency and stability in analyzing sample properties from different modalities for pathologists. Aim: In this work, we propose a computational image translation technique based on deep learning to enable bright-field microscopy contrast using snapshot Stokes images of stained pathological tissue slides. Taking Stokes images as input instead of MM images allows the translated bright-field images to be unaffected by variations of light source and samples. Approach: We adopted CycleGAN as the translation model to avoid requirements on co-registered image pairs in the training. This method can generate images that are equivalent to the bright-field images with different staining styles on the same region. Results: Pathological slices of liver and breast tissues with hematoxylin and eosin staining and lung tissues with two types of immunohistochemistry staining, i.e., thyroid transcription factor-1 and Ki-67, were used to demonstrate the effectiveness of our method. The output results were evaluated by four image quality assessment methods. Conclusions: By comparing the cross-modality translation performance with MM images, we found that the Stokes images, with the advantages of faster acquisition and independence from light intensity and image registration, can be well translated to bright-field images.

Mueller matrix imaging of pathological slides with plastic coverslips.

Mueller matrix microscopy is capable of polarization characterization of pathological samples and polarization imaging based digital pathology. In recent years, hospitals are replacing glass coverslips with plastic coverslips for automatic preparations of dry and clean pathological slides with less slide-sticking and air bubbles. However, plastic coverslips are usually birefringent and introduce polarization artifacts in Mueller matrix imaging. In this study, a spatial frequency based calibration method (SFCM) is used to remove such polarization artifacts. The polarization information of the plastic coverslips and the pathological tissues are separated by the spatial frequency analysis, then the Mueller matrix images of pathological tissues are restored by matrix inversions. By cutting two adjacent lung cancer tissue slides, we prepare paired samples of very similar pathological structures but one with a glass coverslip and the other with a plastic coverslip. Comparisons between Mueller matrix images of the paired samples show that SFCM can effectively remove the artifacts due to plastic coverslip.

Mueller matrix imaging with a spatially modulated polarization light source.

In this paper, we present a Mueller matrix imaging system consisting of a spatially modulated polarization light source (SMPL) and a dual division-of-focal-plane (DoFP) polarimeters as the PSA and 2D detector. The system does not contain moving parts such as a rotating stage, which leads to more robust and reliable operations for applications in hostile settings. By taking Muller matrix images at variable distances between the SMPL and the target, we examine in details errors due to different spatial distributions in angle and intensity of different polarized lights. A calibration method is proposed to reduce such errors introduced by SMPL. The performances of the new imaging technique and the calibration method are tested in Mueller matrix imaging of different samples.

Dual division of focal plane polarimeters-based collinear reflection Mueller matrix fast imaging microscope.

SIGNIFICANCE: Reflection Mueller matrix imaging is suitable for characterizing the microstructure of bulk specimens and probing dynamic processes in living animals, there are always demands for speed and accuracy for such applications to avoid possible artifacts and reveal a sample's intrinsic properties. AIM: To demonstrate a design of collinear reflection Mueller matrix fast imaging microscope based on dual division of focal plane (DoFP) polarimeters (DoFPs-CRMMM) which has high measurement speed and accuracy. APPROACH: In DoFPs-CRMMM, to improve the measurement speed, we applied the dual DoFP polarimeters design on the collinear reflection system for the first time to achieve fast imaging in about 2 s. To improve the measurement accuracy, we improved the double-pass eigenvalue calibration method (dp-ECM) by background light correction, and explored the optimization of the set of reference samples. RESULTS: DoFPs-CRMMM was applied to measure the standard polarization samples and monitor the tissue optical clearing process of an artificial layered bulk tissue. Results show that the system has satisfactory performance which can capture the variation of polarization properties during the dynamic process. CONCLUSIONS: We present the establishment and demo application of DoFPs-CRMMM. The measurement speed can be further accelerated for potential applications in monitoring dynamic processes or living biomedical systems.

Deep learning for denoising in a Mueller matrix microscope.

The Mueller matrix microscope is a powerful tool for characterizing the microstructural features of a complex biological sample. Performance of a Mueller matrix microscope usually relies on two major specifications: measurement accuracy and acquisition time, which may conflict with each other but both contribute to the complexity and expenses of the apparatus. In this paper, we report a learning-based method to improve both specifications of a Mueller matrix microscope using a rotating polarizer and a rotating waveplate polarization state generator. Low noise data from long acquisition time are used as the ground truth. A modified U-Net structured network incorporating channel attention effectively reduces the noise in lower quality Mueller matrix images obtained with much shorter acquisition time. The experimental results show that using high quality Mueller matrix data as ground truth, such a learning-based method can achieve both high measurement accuracy and short acquisition time in polarization imaging.

Rapidly Measuring Scattered Polarization Parameters of the Individual Suspended Particle with Continuously Large Angular Range.

Suspended particles play a vital role in aquatic environments. We propose a method to rapidly measure the scattered polarization parameters of individual suspended particles with continuously large angular range (PCLAR), from 60° to 120° in one shot. A conceptual setup is built to measure PCLAR with 20 kHz; to verify the setup, 10 µm-diameter silica microspheres suspended in water, whose PCLAR are consistent with those simulated by Mie theory, are measured. PCLAR of 6 categories of particles are measured, which enables high-accuracy classification with the help of a convolutional neural network algorithm. PCLAR of different mixtures of Cyclotella stelligera and silica microspheres are measured to successfully identify particulate components. Furthermore, classification ability comparisons of different angular-selection strategies show that PCLAR enables the best classification beyond the single angle, discrete angles and small-ranged angles. Simulated PCLAR of particles with different size, refractive index, and structure show explicit discriminations between them. Inversely, the measured PCLAR are able to estimate the effective size and refractive index of individual Cyclotella cells. Results demonstrate the method's power, which intrinsically takes the advantage of the optical polarization and the angular coverage. Future prototypes based on this concept would be a promising biosensor for particles in environmental monitoring.

Deep learning Mueller matrix feature retrieval from a snapshot Stokes image.

A Mueller matrix (MM) provides a comprehensive representation of the polarization properties of a complex medium and encodes very rich information on the macro- and microstructural features. Histopathological features can be characterized by polarization parameters derived from MM. However, a MM must be derived from at least four Stokes vectors corresponding to four different incident polarization states, which makes the qualities of MM very sensitive to small changes in the imaging system or the sample during the exposures, such as fluctuations in illumination light and co-registration of polarization component images. In this work, we use a deep learning approach to retrieve MM-based specific polarimetry basis parameters (PBPs) from a snapshot Stokes vector. This data post-processing method is capable of eliminating errors introduced by multi-exposure, as well as reducing the imaging time and hardware complexity. It shows the potential for accurate MM imaging on dynamic samples or in unstable environments. The translation model is designed based on generative adversarial network with customized loss functions. The effectiveness of the approach was demonstrated on liver and breast tissue slices and blood smears. Finally, we evaluated the performance by quantitative similarity assessment methods in both pixel and image levels.

Computational image translation from Mueller matrix polarimetry to bright-field microscopy.

Mueller matrix (MM) polarimetry can provide comprehensive information about the polarization properties that are closely related to the microstructural features and has demonstrated its potential in biomedical studies and clinical practices, and bright-field microscopy is widely used in pathological diagnosis as the golden standard. In this work, we improve the throughput of MM microscopy by learning a statistical transformation between these two imaging systems based on deep learning. Using this approach, the MM microscope can generate an image that is equivalent to a bright-field microscope image of the matching field of view. We add new transformative capability to the existing MM imaging system without requiring extra hardware. The translation model is based on conditional generative adversarial network with customized loss functions. We demonstrated the effectiveness of our approach on liver and breast tissues and evaluated the performance by four quantitative similarity assessment methods in pixel, image and distribution levels, respectively.

Geometric optimization method for a polarization state generator of a Mueller matrix microscope.

We propose a geometric optimization method combined with the Coulombic energy indicator that can uniformly distribute N polarization states on the Poincaré sphere. Based on this method, we investigate the optimal frames of a rotating polarizer and rotating quarter-wave plate (RPRQ)-based polarization state generator (PSG) at different numbers of modulations. We use the PSG on a dual DoFP polarimeter-based Mueller matrix microscope to measure standard samples and pathological sections for testing the performance of an optimized RPRQ. The experimental results show that this method can effectively restrain noise and improve measurement accuracy.

Fast Mueller matrix microscope based on dual DoFP polarimeters.

In this Letter, we report a dual division of focal plane (DoFP) polarimeters-based full Mueller matrix microscope (DoFPs-MMM) for fast polarization imaging. Both acquisition speed and measurement accuracy are improved compared with those of a Mueller matrix microscope based on dual rotating retarders. Then, the system is applied to probe the polarization properties of a red blood cells smear. The experimental results show that a DoFPs-MMM has the potential to be a powerful tool for probing dynamic processes in living cells in future studies.