Classification of marine microalgae using low-resolution Mueller matrix images and convolutional neural network

Imaging and Non-imaging Analytical Techniques Used for Drug Nanosizing and their Patents: An Overview

BACKGROUND

Nanosizing is widely recognized as an effective technique for improving the solubility, dissolution rate, onset of action, and bioavailability of poorly water-soluble drugs. To control the execution and behavior of the output product, more advanced and valuable analytical techniques are required.

OBJECTIVE

The primary intent of this review manuscript was to furnish the understanding of imaging and non-imaging techniques related to nanosizing analysis by focusing on related patents. In addition, the study also aimed to collect and illustrate the information on various classical (laser diffractometry, photon correlation spectroscopy, zeta potential, laser Doppler electrophoresis, X-ray diffractometry, differential scanning calorimeter, scanning electron microscopy, transmission electron microscopy), new, and advanced analytical techniques (improved dynamic light scattering method, Brunauer-Emmett- Teller method, ultrasonic attenuation, biosensor), as well as commercial techniques, like inductively coupled plasma mass spectroscopy, aerodynamic particle sizer, scanning mobility particle sizer, and matrix- assisted laser desorption/ionization mass spectroscopy, which all relate to nano-sized particles.

METHODS

The present manuscript has taken a fresh look at the various aspects of the analytical techniques utilized in the process of nanosizing, and has achieved this through the analysis of a wide range of peer-reviewed literature. All summarized literature studies provide the information that can meet the basic needs of nanotechnology.

RESULTS

A variety of analytical techniques related to the nanosizing process have already been established and have great potential to weed out several issues. However, the current scenarios require more relevant, accurate, and advanced analytical techniques that can minimize the time and deviations associated with different instrumental and process parameters. To meet this requirement, some new and more advanced analytical techniques have recently been discovered, like ultrasonic attenuation technique, BET technique, biosensors, etc. Conclusion: The present overview certifies the significance of different analytical techniques utilized in the nanosizing process. The overview also provides information on various patents related to sophisticated analytical tools that can meet the needs of such an advanced field. The data show that the nanotechnology field will flourish in the coming future.

Recognition of microplastics suspended in seawater via refractive index by Mueller matrix polarimetry

Microplastics have become the marine pollution posing a human health risk, but they are difficult to be detected and recognized for different materials, irregular shapes, and broad size distributions. Microplastics' refractive index (RI) is related to the materials and can be characterized by the Mueller matrix. In this work, the particles are suspended in water and their Mueller matrices are measured by a particulate Mueller matrix polarimetry setup. Four kinds of spherical particles including microplastics are effectively classified by their Mueller matrices. Moreover, two kinds of common microplastics with broad size distributions, irregular shapes, and random orientations are also well recognized by the Mueller matrix. These results imply that RI plays a vital role in the recognition of microplastics suspended in water. By using the Mie theory and discrete dipole approximation simulation, the discussions explain in physics origin how RI affects Mueller matrix coupling with size and structure, and give some decoupling methods. Results in this work help advance future tools to in situ recognize the microplastics in seawater.

Dual Polarization Modality Fusion Network for Assisting Pathological Diagnosis

Polarization imaging is sensitive to sub-wavelength microstructures of various cancer tissues, providing abundant optical characteristics and microstructure information of complex pathological specimens. However, how to reasonably utilize polarization information to strengthen pathological diagnosis ability remains a challenging issue. In order to take full advantage of pathological image information and polarization features of samples, we propose a dual polarization modality fusion network (DPMFNet), which consists of a multi-stream CNN structure and a switched attention fusion module for complementarily aggregating the features from different modality images. Our proposed switched attention mechanism could obtain the joint feature embeddings by switching the attention map of different modality images to improve their semantic relatedness. By including a dual-polarization contrastive training scheme, our method can synthesize and align the interaction and representation of two polarization features. Experimental evaluations on three cancer datasets show the superiority of our method in assisting pathological diagnosis, especially in small datasets and low imaging resolution cases. Grad-CAM visualizes the important regions of the pathological images and the polarization images, indicating that the two modalities play different roles and allow us to give insightful corresponding explanations and analysis on cancer diagnosis conducted by the DPMFNet. This technique has potential to facilitate the performance of pathological aided diagnosis and broaden the current digital pathology boundary based on pathological image features.

Combined Mueller matrix imaging and artificial intelligence classification framework for Hepatitis B detection

SIGNIFICANCE

The combination of polarized imaging with artificial intelligence (AI) technology has provided a powerful tool for performing an objective and precise diagnosis in medicine.

AIM

An approach is proposed for the detection of hepatitis B (HB) virus using a combined Mueller matrix imaging technique and deep learning method.

APPROACH

In the proposed approach, Mueller matrix imaging polarimetry is applied to obtain 4 × 4 Mueller matrix images of 138 HBsAg-containing (positive) serum samples and 136 HBsAg-free (negative) serum samples. The kernel estimation density results show that, of the 16 Mueller matrix elements, elements M 22 and M 33 provide the best discriminatory power between the positive and negative samples.

RESULTS

As a result, M 22 and M 33 are taken as the inputs to five different deep learning models: Xception, VGG16, VGG19, ResNet 50, and ResNet150. It is shown that the optimal classification accuracy (94.5%) is obtained using the VGG19 model with element M 22 as the input.

CONCLUSIONS

Overall, the results confirm that the proposed hybrid Mueller matrix imaging and AI framework provides a simple and effective approach for HB virus detection.

Profile probing of suspended particles in water by Stokes vector polarimetry

Suspended particles are the important components of natural water. In this paper, a method based on polarized light scattering is proposed for profile probing of the particulate components in water. The profile probing is achieved by a polarized light sheet illuminating the suspension and the Stokes vector imaging system at a 120° backscattering angle, receiving the scattered light of the particles in the scattering volume. Each Stokes vector image (SVI) includes hundreds of star-studded particles whose Stokes vectors are used to retrieve the numbers of each particulate component in water. Experiments of typical particles are conducted. The classifications of these particles powered by the convolutional neural network (CNN) are demonstrated. The particulate components in mixed samples are successfully recognized and quantitatively compared. Considering at least 10 SVIs every second, the concentrations of each particulate component in water are effectively evaluated. The concept of profile probing the particulate components in water is proved to be powerful, by which we can measure up to almost 8000 particles per second. These results encourage the development of in-situ tools with this concept for particle profiling in future field surveying.

Detection Methods of Nanoparticles Synthesized by Gas-Phase Method: A Review

The detection of nanoparticles is the basis of the study of synthesis mechanism, active regulation of the synthesis process, and the study of nanoparticle properties after synthesis. It is significantly meaningful to the academia and engineering industry. Although there are many relevant detection methods at present, each method has its own advantages and disadvantages, and their measurement quantity and application conditions are also different. There is a lack of unified sorting and generalization. In this paper, the significance of detection of nanoparticles synthesized by a gas-phase method is introduced, the development of detection technology is reviewed, and the future is prospected. It is hoped that this paper will provide a reference for the detection of nanoparticles under various conditions and for the development of new detection methods.

A Method for Optimal Detection of Lung Cancer Based on Deep Learning Optimized by Marine Predators Algorithm

Lung cancer is the uncontrolled growth of cells in the lung that are made up of two spongy organs located in the chest. These cells may penetrate outside the lungs in a process called metastasis and spread to tissues and organs in the body. In this paper, using image processing, deep learning, and metaheuristic, an optimal methodology is proposed for early detection of this cancer. Here, we design a new convolutional neural network for this purpose. Marine predators algorithm is also used for optimal arrangement and better network accuracy. The method finally applied to RIDER dataset, and the results are compared with some pretrained deep networks, including CNN ResNet-18, GoogLeNet, AlexNet, and VGG-19. Final results showed higher results of the proposed method toward the compared techniques. The results showed that the proposed MPA-based method with 93.4% accuracy, 98.4% sensitivity, and 97.1% specificity provides the highest efficiency with the least error (1.6) toward the other state of the art methods.

MuellerNet: a hybrid 3D-2D CNN for cell classification with Mueller matrix images

Different from conventional microimaging techniques, polarization imaging can generate multiple polarization images in a single perspective by changing the polarization angle. However, how to efficiently fuse the information in these multiple polarization images by a convolutional neural network (CNN) is still a challenging problem. In this paper, we propose a hybrid 3D-2D convolutional neural network called MuellerNet, to classify biological cells with Mueller matrix images (MMIs). The MuellerNet includes a normal stream and a polarimetric stream, in which the first Mueller matrix image is taken as the input of normal stream, and the rest MMIs are stacked to form the input of a polarimetric stream. The normal stream is mainly constructed with a backbone network and, in the polarimetric stream, the attention mechanism is used to adaptively assign weights to different convolutional maps. To improve the network's discrimination, a loss function is introduced to simultaneously optimize parameters of the two streams. Two Mueller matrix image datasets are built, which include four types of breast cancer cells and three types of algal cells, respectively. Experiments are conducted on these two datasets with many well-known and recent networks. Results show that the proposed network efficiently improves the classification accuracy and helps to find discriminative features in MMIs.