The application and comparison of U-Net convolutional neural network architectures is proposed in this work to guarantee a fast and accurate convergence of inverse scattering problems solved by Born iterative method, even in the presence of strong scatterers. Starting from a preliminary configuration proposed by the authors in some recent papers, two variants are introduced and discussed to significantly reduce the computational cost, while guaranteeing convergence with very high accuracy in the dielectric profiles reconstruction when considering strong scattereres, such as tumors, thus working as a regularization process to mitigate the induced non-linearity. As a further enhancement, a novel approach is introduced which integrates U-Net and Resnet models to realize a segmentation process, thus leading to the effective feature extraction and the accurate identification of anomalies within healthy tissue. Numerical assessments on a variety of breast models including abnormal lesions are discussed to successfully validate the proposed machine learning approach, through the adoption of properly defined evaluation metrics.
In this work, a novel technique is proposed that combines the Born iterative method, based on a quadratic programming approach, with convolutional neural networks to solve the ill-framed inverse problem coming from microwave imaging formulation in breast cancer detection. The aim is to accurately recover the permittivity of breast phantoms, these typically being strong dielectric scatterers, from the measured scattering data. Several tests were carried out, using a circular imaging configuration and breast models, to evaluate the performance of the proposed scheme, showing that the application of convolutional neural networks allows clinicians to considerably reduce the reconstruction time with an accuracy that exceeds 90% in all the performed validations.
An improved machine learning approach is presented in this paper to guarantee the fast convergence of the Born Iterative Method, even in the presence of strong scatterers, by assuming a single operating frequency and a reduced number of antennas in the scattering setup. The initial estimation of the dielectric profile, provided by the Born Iterative Method, was processed by a specific convolutional neural network to improve the reconstruction using a fast machine learning approach. To ensure rapid convergence, a proper choice of the initial guess in terms of the minimum permittivity value over the entire domain was also made. Numerical validations on realistic breast phantoms were illustrated, demonstrating an average error of 2.4% and an accuracy greater than 96% for all considered tests, even when considering a single operating frequency and a reduced amount of training data.
COVID-19 is an infectious disease mainly transmitted through aerosol particles. Physical distancing can significantly reduce airborne transmission at a short range, but it is not a sufficient measure to avoid contagion. In recent months, health authorities have identified indoor spaces as possible sources of infection, mainly due to poor ventilation, making it necessary to take measures to improve indoor air quality. In this work, an accurate model for COVID-19 contagion risk estimation based on the Wells–Riley probabilistic approach for indoor environments is proposed and implemented as an Android mobile App. The implemented algorithm takes into account all relevant parameters, such as environmental conditions, age, kind of activities, and ventilation conditions, influencing the risk of contagion to provide the real-time probability of contagion with respect to the permanence time, the maximum allowed number of people for the specified area, the expected number of COVID-19 cases, and the required number of Air Changes per Hour. Alerts are provided to the user in the case of a high probability of contagion and CO2 concentration. Additionally, the app exploits a Bluetooth signal to estimate the distance to other devices, allowing the regulation of social distance between people. The results from the application of the model are provided and discussed for different scenarios, such as offices, restaurants, classrooms, and libraries, thus proving the effectiveness of the proposed tool, helping to reduce the spread of the virus still affecting the world population.
This study discusses a new technique based on the Born iterative method, with a quadratic programming approach, which uses convolutional neural networks to solve microwave inverse scattering problems for breast cancer detection. The ultimate goal is to reconstruct the permittivity of breast tissues from the measured scattering values, which poses a great challenge because these tissues are characterized by strong dielectric scatterers. To evaluate the performance of the proposed method, simulations were performed for two breast phantom models with tumors in different positions. The results obtained are very good and show that the application of Convolutional Neural Networks helps to significantly reduce the reconstruction error.
Complex-valued neural networks have emerged as an effective instrument in image reconstruction, exhibiting significant advancements compared to conventional techniques. This study introduces an innovative methodology to tackle the difficulties related to image reconstruction within medical microwave imaging. Initially, in the estimation phase, the proposed methodology integrates the Born iterative method with quadratic programming. Subsequently, in the refinement stage, the study explores the application of complex-valued neural networks to enhance the quality of reconstructions. The research emphasizes distinct complex-valued neural network architectures, namely, CV-UNET, CV-CNN, CV-MLP, and their corresponding performances. CV-UNET stands out as the best architecture, surpassing conventional methods and the other complex-valued neural networks variants. The complex-valued neural network improves the fidelity of reconstructions and simplifies the procedure by obviating the need for multiple training steps, a common prerequisite in real-valued neural networks.