The Use of Artificial Intelligence in the Liver Histopathology Field: A Systematic Review.

Digital pathology (DP) has begun to play a key role in the evaluation of liver specimens. Recent studies have shown that a workflow that combines DP and artificial intelligence (AI) applied to histopathology has potential value in supporting the diagnosis, treatment evaluation, and prognosis prediction of liver diseases. Here, we provide a systematic review of the use of this workflow in the field of hepatology. Based on the PRISMA 2020 criteria, a search of the PubMed, SCOPUS, and Embase electronic databases was conducted, applying inclusion/exclusion filters. The articles were evaluated by two independent reviewers, who extracted the specifications and objectives of each study, the AI tools used, and the results obtained. From the 266 initial records identified, 25 eligible studies were selected, mainly conducted on human liver tissues. Most of the studies were performed using whole-slide imaging systems for imaging acquisition and applying different machine learning and deep learning methods for image pre-processing, segmentation, feature extractions, and classification. Of note, most of the studies selected demonstrated good performance as classifiers of liver histological images compared to pathologist annotations. Promising results to date bode well for the not-too-distant inclusion of these techniques in clinical practice.

Optimized Leaky-Wave Antenna for Hyperthermia in Biological Tissue Theoretical Model.

In this paper, we exploit the enhanced penetration reachable through inhomogeneous waves to induce hyperthermia in biological tissues. We will present a leaky-wave antenna inspired by the Menzel antenna which has been shortened through opportune design and optimizations and that has been designed to optimize the penetration at the interface with the skin, allowing penetration in the skin layer at a constant temperature, and enhanced penetration in the overall structure considered. Past papers both numerically and analytically demonstrated the possibility of reducing the attenuation that the electromagnetic waves are subject to when travelling inside a lossy medium by using inhomogeneous waves. In those papers, a structure (the leaky-wave antenna) is shown to allow the effect, but such a radiator suffers from low efficiency. Also, at the frequencies that are most used for hyperthermia application, a classical leaky-wave antenna would be too long; here is where the idea of the shortened leaky-wave arises. To numerically analyze the penetration in biological tissues, this paper considers a numerical prototype of a sample of flesh, composed of superficial skin layers, followed by fat and an undefined layer of muscles.

Melanoma Detection by Non-Specialists: An Untapped Potential for Triage?

INTRODUCTION: The incidence of melanoma increased considerably in recent decades, representing a significant public health problem. We aimed to evaluate the ability of non-specialists for the preliminary screening of skin lesions to identify melanoma-suspect lesions. MATERIALS AND METHODS: A medical student and a dermatologist specialist examined the total body scans of 50 patients. RESULTS: The agreement between the expert and the non-specialist was 87.75% (κ = 0.65) regarding the assessment of clinical significance. The four parameters of the ABCD rule were evaluated on the 129 lesions rated as clinically significant by both observers. Asymmetry was evaluated similarly in 79.9% (κ = 0.59), irregular borders in 74.4% (κ = 0.50), color in 81.4% (κ = 0.57), and diameter in 89.9% (κ = 0.77) of the cases. The concordance of the two groups was 96.9% (κ = 0.83) in the case of the detection of the Ugly Duckling Sign. CONCLUSIONS: Although the involvement of GPs is part of routine care worldwide, emphasizing the importance of educating medical students and general practitioners is crucial, as many European countries lack structured melanoma screening training programs targeting non-dermatologists.

Deep Learning Hybrid Techniques for Brain Tumor Segmentation.

Medical images play an important role in medical diagnosis and treatment. Oncologists analyze images to determine the different characteristics of deadly diseases, plan the therapy, and observe the evolution of the disease. The objective of this paper is to propose a method for the detection of brain tumors. Brain tumors are identified from Magnetic Resonance (MR) images by performing suitable segmentation procedures. The latest technical literature concerning radiographic images of the brain shows that deep learning methods can be implemented to extract specific features of brain tumors, aiding clinical diagnosis. For this reason, most data scientists and AI researchers work on Machine Learning methods for designing automatic screening procedures. Indeed, an automated method would result in quicker segmentation findings, providing a robust output with respect to possible differences in data sources, mostly due to different procedures in data recording and storing, resulting in a more consistent identification of brain tumors. To improve the performance of the segmentation procedure, new architectures are proposed and tested in this paper. We propose deep neural networks for the detection of brain tumors, trained on the MRI scans of patients' brains. The proposed architectures are based on convolutional neural networks and inception modules for brain tumor segmentation. A comparison of these proposed architectures with the baseline reference ones shows very interesting results. MI-Unet showed a performance increase in comparison to baseline Unet architecture by 7.5% in dice score, 23.91% insensitivity, and 7.09% in specificity. Depth-wise separable MI-Unet showed a performance increase by 10.83% in dice score, 2.97% in sensitivity, and 12.72% in specificity as compared to the baseline Unet architecture. Hybrid Unet architecture achieved performance improvement of 9.71% in dice score, 3.56% in sensitivity, and 12.6% in specificity. Whereas the depth-wise separable hybrid Unet architecture outperformed the baseline architecture by 15.45% in dice score, 20.56% in sensitivity, and 12.22% in specificity.

Artificial Intelligence for Thyroid Nodule Characterization: Where Are We Standing?

Machine learning (ML) is an interdisciplinary sector in the subset of artificial intelligence (AI) that creates systems to set up logical connections using algorithms, and thus offers predictions for complex data analysis. In the present review, an up-to-date summary of the current state of the art regarding ML and AI implementation for thyroid nodule ultrasound characterization and cancer is provided, highlighting controversies over AI application as well as possible benefits of ML, such as, for example, training purposes. There is evidence that AI increases diagnostic accuracy and significantly limits inter-observer variability by using standardized mathematical algorithms. It could also be of aid in practice settings with limited sub-specialty expertise, offering a second opinion by means of radiomics and computer-assisted diagnosis. The introduction of AI represents a revolutionary event in thyroid nodule evaluation, but key issues for further implementation include integration with radiologist expertise, impact on workflow and efficiency, and performance monitoring.

Verification of the electromagnetic deep-penetration effect in the real world.

The deep penetration of electromagnetic waves into lossy media can be obtained by properly generating inhomogeneous waves. In this work, for the very first time, we demonstrate the physical implementation and the practical relevance of this phenomenon. A thorough numerical investigation of the deep-penetration effects has been performed by designing and comparing three distinct practical radiators, emitting either homogeneous or inhomogeneous waves. As concerns the latter kind, a typical Menzel microstrip antenna is first used to radiate improper leaky waves. Then, a completely new approach based on an optimized 3-D horn TEM antenna applied to a lossy prism is described, which may find applications even at optical frequencies. The effectiveness of the proposed radiators is measured using different algorithms to consider distinct aspects of the propagation in lossy media. We finally demonstrate that the deep penetration is possible, by extending the ideal and theoretical evidence to practical relevance, and discuss both achievements and limits obtained through numerical simulations on the designed antennas.

Scattering of Light from the Systemic Circulatory System.

There are many factors of methodological origin that influence the measurement of optical properties of the entire circulatory system which consists of blood as the basic component. The basic idea of this review article is to provide the optical properties of the circulatory system with all those factors of influence that have been employed in biomedical optics for different applications. We begin with the available optical properties, i.e., absorption, scattering and, reduced scattering coefficient, in general for any tissue inside the human body and prominent scattering theories (e.g., light, X-rays, neutrons) that are helpful in this regard. We have reviewed and compiled already available formulas and their respective available data for different human tissues for these optical properties. Then we have descended to the blood composition and to different scattering techniques available in the literature to study scattering and light propagation inside blood. We have reviewed both computational and theoretical scattering techniques.

Introduction to electromagnetic scattering, part II: tutorial.

In this paper, some generalizations of electromagnetic scattering problems by elementary shapes are presented. In particular, the aim of the paper is to provide solutions to the scattering problem by multiple objects with simple shapes, either in concentric configuration or arbitrarily distributed in the space. The vector harmonics, representing the fields, and their properties are applied in order to solve five different problems: the electromagnetic scattering by an infinitely long circular stratified cylinder, by a multilayered sphere, by an ensemble of parallel cylinders, by an ensemble of multi-spheres, and ultimately by a sphere embedded in a circular cylinder. Numerical results in particularly important configurations are shown.

Numerical simulation of the blood oxygenation level-dependent functional magnetic resonance signal using finite element method.

Since the introduction of functional magnetic resonance imaging (fMRI), several computational approaches have been developed to examine the effect of the morphology and arrangement of blood vessels on the blood oxygenation-level dependent (BOLD) signal in the brain. In the present work, we implemented the original Ogawa's model using a numerical simulation based on the finite element method (FEM) instead of the analytical models. In literature, there are different works using analytical methods to analyse the transverse relaxation rate ( R2∗ ), which BOLD signal is related to, modelling the vascular system with simple and canonical geometries such as an infinite cylinder model (ICM) or a set of cylinders. We applied the numerical simulation to the extravascular BOLD signal as a function of angular vessel distribution (perpendicular vs parallel to the static magnetic field) relevant for anatomical districts characterized by geometrical symmetries, such as spinal cord. Numerical simulations confirmed analytical results for the canonical ICM. Moreover, the perturbation to the magnetic field induced by blood deoxyhaemoglobin, as quantified assuming Brownian diffusion of water molecules around the vessel, revealed that vessels contribute the most to the variation of the R2∗ when they are preferentially perpendicular to the external magnetic field, regardless of their size. Our results indicate that the numerical simulation method is sensitive to the effects of different vascular geometry. This work highlights the opportunity to extend R2∗ simulations to realistic models of vasculature based on high-resolution anatomical images.

Cancer Diagnosis Using Deep Learning: A Bibliographic Review.

In this paper, we first describe the basics of the field of cancer diagnosis, which includes steps of cancer diagnosis followed by the typical classification methods used by doctors, providing a historical idea of cancer classification techniques to the readers. These methods include Asymmetry, Border, Color and Diameter (ABCD) method, seven-point detection method, Menzies method, and pattern analysis. They are used regularly by doctors for cancer diagnosis, although they are not considered very efficient for obtaining better performance. Moreover, considering all types of audience, the basic evaluation criteria are also discussed. The criteria include the receiver operating characteristic curve (ROC curve), Area under the ROC curve (AUC), F1 score, accuracy, specificity, sensitivity, precision, dice-coefficient, average accuracy, and Jaccard index. Previously used methods are considered inefficient, asking for better and smarter methods for cancer diagnosis. Artificial intelligence and cancer diagnosis are gaining attention as a way to define better diagnostic tools. In particular, deep neural networks can be successfully used for intelligent image analysis. The basic framework of how this machine learning works on medical imaging is provided in this study, i.e., pre-processing, image segmentation and post-processing. The second part of this manuscript describes the different deep learning techniques, such as convolutional neural networks (CNNs), generative adversarial models (GANs), deep autoencoders (DANs), restricted Boltzmann's machine (RBM), stacked autoencoders (SAE), convolutional autoencoders (CAE), recurrent neural networks (RNNs), long short-term memory (LTSM), multi-scale convolutional neural network (M-CNN), multi-instance learning convolutional neural network (MIL-CNN). For each technique, we provide Python codes, to allow interested readers to experiment with the cited algorithms on their own diagnostic problems. The third part of this manuscript compiles the successfully applied deep learning models for different types of cancers. Considering the length of the manuscript, we restrict ourselves to the discussion of breast cancer, lung cancer, brain cancer, and skin cancer. The purpose of this bibliographic review is to provide researchers opting to work in implementing deep learning and artificial neural networks for cancer diagnosis a knowledge from scratch of the state-of-the-art achievements.

An Analytical Study of Electromagnetic Deep Penetration Conditions and Implications in Lossy Media through Inhomogeneous Waves.

This paper illustrates how the penetration of electromagnetic waves in lossy media strongly depends on the waveform and not only on the media involved. In particular, the so-called inhomogeneous plane waves are compared against homogeneous plane waves illustrating how the first ones can generate deep penetration effects. Moreover, the paper provides examples showing how such waves may be practically generated. The approach taken here is analytical and it concentrates on the deep penetration conditions obtained by means of incident inhomogeneous plane waves incoming from a lossless medium and impinging on a lossy medium. Both conditions and constraints that the waveforms need to possess to achieve deep penetration are analysed. Some results are finally validated through numerical computations. The theory presented here is of interest in view of a practical implementation of the deep penetration effect.

Numerical analysis of electromagnetic interactions by a cell during the mitosis phases.

In this work, a numerical study of the interaction of an electromagnetic field with a biological cell in the different phases of mitosis is presented. In particular, the validity of the single-shell model during mitosis from a geometrical viewpoint has been considered. We consider the Jurkat cell for our analysis and model the content of the cell, composed by the nucleus and different organelles immersed in cytoplasm and delimited by the cell membrane, as a single material with appropriate electromagnetic properties derived by making use of effective medium approximation methods, i.e., we used a quasistatic approach to deal with the cell. To validate the model, a comparison between the original and 2 approximated geometry models is made and the results are in very good accordance. Subsequently, an analysis of the scattered field indicates that the most sensitive component to the mitosis phase is the one parallel to the segment joining the centers of the cells.

Polarization-maintaining reflection-mode THz time-domain spectroscopy of a polyimide based ultra-thin narrow-band metamaterial absorber.

This paper reports the design, the microfabrication and the experimental characterization of an ultra-thin narrow-band metamaterial absorber at terahertz frequencies. The metamaterial device is composed of a highly flexible polyimide spacer included between a top electric ring resonator with a four-fold rotational symmetry and a bottom ground plane that avoids misalignment problems. Its performance has been experimentally demonstrated by a custom polarization-maintaining reflection-mode terahertz time-domain spectroscopy system properly designed in order to reach a collimated configuration of the terahertz beam. The dependence of the spectral characteristics of this metamaterial absorber has been evaluated on the azimuthal angle under oblique incidence. The obtained absorbance levels are comprised between 67% and 74% at 1.092 THz and the polarization insensitivity has been verified in transverse electric polarization. This offers potential prospects in terahertz imaging, in terahertz stealth technology, in substance identification, and in non-planar applications. The proposed compact experimental set-up can be applied to investigate arbitrary polarization-sensitive terahertz devices under oblique incidence, allowing for a wide reproducibility of the measurements.

Introduction to electromagnetic scattering: tutorial.

In this paper, an introduction to electromagnetic scattering is presented. We introduce the basic concepts needed to face a scattering problem, including the scattering, absorption, and extinction cross sections. We define the vector harmonics and we present some of their properties. Finally, we tackle the two canonical problems of the scattering by an infinitely long circular cylinder, and by a sphere, showing that the introduction of the vector wave function makes the imposition and solution of the boundary conditions particularly simple.

Electromagnetic inhomogeneous waves at planar boundaries: tutorial.

In this review paper, we summarize the fundamental properties of inhomogeneous waves at the planar interface between two media. We point out the main differences between the wave types: lateral waves, surface waves, and leaky waves. We analyze each kind of inhomogeneous wave, giving a quasi-optical description and explaining the physical origin of some of their properties.

In silico validation procedure for cell volume fraction estimation through dielectric spectroscopy.

Dielectric spectroscopy has proved to be a good tool for analyzing the passive electrical properties of biological tissues as well as those of inhomogeneous materials. This technique promises to be a valid alternative to the classical ones based on metabolites to monitor the growth and cell volume fraction of cell cultures in a simple and minimally invasive way. In order to obtain an accurate estimation of the cell volume fraction as a function of the permittivity of the suspension, a simple in silico procedure is proposed. The procedure is designed to perform homogenization from the micro-scale to the macro-scale using simple analytical models and simulation setups hypothesizing the properties of diluted suspension (cell volume fraction less than 0.2). Results obtained show the possibility to overcome some trouble involving the analytical treatment of the cellular shape by considering a sphere with the same permittivity in the quantitative analysis of the cell volume fraction. The entire study is based on computer simulations performed in order to verify the correctness of the procedure. Obtained data are used in a cell volume fraction estimation scenario to show the effectiveness of the procedure.

Analysis of the polarizability of an array of spherical metallic inclusions in a dielectric host sphere.

The polarizability of an array of metallic spheres embedded in a dielectric host sphere is obtained by means of a quasi-static analysis of the electromagnetic interaction. The proposed model is validated through comparisons with the results obtained with software based on the finite element method. A parametric study of the polarizability as a function of the number of inclusions, their radii, and their positions is presented. An analysis of the plasmon resonances of the particle as a function of the same parameters is performed.

Generalized image principle for cylindrical waves.

In this Letter, we analyze the reflection of cylindrical waves (CWs) at planar interfaces. We consider the reflected CW proposed in the literature as a spectral integral. We present a Laurent series expansion of the Fresnel coefficient convergent on the whole real axis and we use it to solve analytically the reflected-wave integral. We found a solution that involves both Bessel functions and Anger-Weber functions, i.e., solutions of both the homogeneous and inhomogeneous Bessel differential equations. We compare the analytical solution with the numerical results obtained with a quadrature formula presented in the literature. Moreover, we present a physical interpretation that connects our solution to the image principle.

Electromagnetic interaction with two eccentric spheres.

In this paper, we consider the interaction of an electromagnetic field with two eccentric spheres. We propose a quasi-static approach in order to calculate the scattered field and the polarizability and the effective permittivity of the eccentric spheres. We analyze the behavior of the scattering parameters as a function of the dimension and position of the spherical inclusions. Moreover, we consider the case of plasmonic spheres and study the behavior of the plasmon resonances for different reciprocal positions of the two spheres.

Electromagnetic scattering by a circular cylinder buried below a slightly rough Gaussian surface.

A two-dimensional beam is scattered by a cylinder buried below a slightly rough surface. The cylindrical wave approach is applied, i.e., cylindrical waves are employed as basis functions of the fields scattered by the cylinder. Moreover, a spectral representation of both the incident field and the cylindrical waves is used. Rough surface deviation is coped with by the first-order small perturbation method. Therefore, to a zeroth-order solution relevant to scattering in the case of a flat surface, a first-order approximation is superimposed. The theoretical approach has been implemented for a periodic surface with Gaussian roughness spectrum.