Explore the vast range of titles available.

A non-uniform antenna array is proposed to enhance the accuracy of medical microwave imaging systems by increasing the amount of useful information captured about the imaged domain without increasing the number of antennas. These systems have so far been using uniform antenna arrays, which lead to highly correlated signals, limiting the amount of imaging information and adversely affecting diagnostic accuracy. In the proposed non-uniform antenna array method, the optimal number and positions of antennas are calculated with the aim of enhancing spatial diversity and reducing information redundancy. The mutual information coefficient is used as a metric to evaluate and minimize redundancy between received signals. A microwave head imaging system is used to verify the proposed approach. The results of the investigated scenarios show that using a non-uniform antenna configuration outperforms a uniform setup in imaging accuracy and clarity, when using the same number of antennas. Moreover, the reconstructed images demonstrate that using an optimized non-uniform antenna array with fewer elements can outperform a uniform array with more elements in terms of localization accuracy and image quality. The proposed approach improves imaging performance and reduces system complexity, cost, and power consumption, making it a practical solution for real-world biomedical imaging applications.

Synthetic microwave focusing methods have been widely adopted in qualitative medical imaging to detect and localize anomalies based on their electromagnetic scattering signatures. This paper discusses the principles, challenges, and limitations of synthetic microwave-focusing techniques in medical applications. It is shown that the various focusing techniques, including time reversal, confocal imaging, and delay-and-sum, are all based on the scalar solution of the electromagnetic scattering problem, assuming the imaged object, i.e., the tissue or object, is linear, reciprocal, and time-invariant. They all aim to generate a qualitative image, revealing any strong scatterer within the imaged domain. The differences among these techniques lie only in the assumptions made to derive the solution and create an image of the relevant tissue or object. To get a fast solution using limited computational resources, those methods assume the tissue is homogeneous and non-dispersive, and thus, a simplified far-field Green’s function is used. Some focusing methods compensate for dispersive effects and attenuation in lossy tissues. Other approaches replace the simplified Green’s function with more representative functions. While these focusing techniques offer benefits like speed and low computational requirements, they face significant ongoing challenges in real-life applications due to their oversimplified linear solutions to the complex problem of non-linear medical microwave imaging. This paper discusses these challenges and potential solutions.

Estimating blood glucose levels (BGLs) noninvasively is a rapidly advancing field driven by the need for effective and painless glucose monitoring solutions for diabetic patients. This study explores deep learning (DL) models applied to noninvasive techniques for accurate BGL estimation. Thermal images were collected for Type I diabetes after confirming BGLs using a glucometer. DL techniques were then employed to classify the thermal images into three BGL classes (low, high, and normal). DarkNet and ShuffleNet convolutional neural networks (CNNs) are used to classify the thermal image and get the best performance, with an overall accuracy of 98% for DarkNet and 100% for ShuffleNet CNN.