This paper presents the detection of brain tumors through the YOLOv3 deep neural network model in a portable electromagnetic (EM) imaging system. YOLOv3 is a popular object detection model with high accuracy and improved computational speed. Initially, the scattering parameters are collected from the nine-antenna array setup with a tissue-mimicking head phantom, where one antenna acts as a transmitter and the other eight antennas act as receivers. The images are then reconstructed from the post-processed scattering parameters by applying the modified delay-multiply-and-sum algorithm that contains 416×416 pixels. Fifty sample images are collected from the different head regions through the EM imaging system. The images are later augmented to generate a final image data set for training, validation, and testing, where the data set contains 1000 images, including fifty samples with a single and double tumor. 80% of the images are utilized for training the network, whereas 10% are used for validation, and the rest 10% are utilized for testing purposes. The detection performance is investigated with the different image data sets. The achieved detection accuracy and F1 scores are 95.62% and 94.50%, respectively, which ensure better detection accuracy. The training accuracy and validation losses are 96.74% and 9.21%, respectively. The tumor detection with its location in different cases from the testing images is evaluated through YOLOv3, which demonstrates its potential in the portable electromagnetic head imaging system.

In this paper, a compact planar ultrawideband (UWB) antenna and an antenna array setup for microwave breast imaging are presented. The proposed antenna is constructed with a slotted semicircular-shaped patch and partial trapezoidal ground. It is compact in dimension: 0.30λ × 0.31λ × 0.011λ, where λ is the wavelength of the lowest operating frequency. For design purposes, several parameters are assumed and optimized to achieve better performance. The prototype is applied in the breast imaging scheme over the UWB frequency range 3.10–10.60 GHz. However, the antenna achieves an operating bandwidth of 8.70 GHz (2.30–11.00 GHz) for the reflection coefficient under–10 dB with decent impedance matching, 5.80 dBi of maximum gain with steady radiation pattern. The antenna provides a fidelity factor (FF) of 82% and 81% for face-to-face and side-by-side setups, respectively, which specifies the directionality and minor variation of the received pulses. The antenna is fabricated and measured to evaluate the antenna characteristics. A 16-antenna array-based configuration is considered to measure the backscattering signal of the breast phantom where one antenna acts as transmitter, and 15 of them receive the scattered signals. The data is taken in both the configuration of the phantom with and without the tumor inside. Later, the Iteratively Corrected Delay and Sum (IC–DAS) image reconstructed algorithm was used to identify the tumor in the breast phantom. Finally, the reconstructed images from the analysis and processing of the backscattering signal by the algorithm are illustrated to verify the imaging performance.

A wideband grounded coplanar waveguide (GCPW) based patch antenna for head imaging is discussed in this paper. The proposed antenna is integrated with a slotted inverted delta-shaped main radiating patch, coplanar waveguide elements, and slotted partial ground. The prime objective of usages of coplanar waveguide elements is enhancing antenna effectiveness. In the prototype, examined different categories of slots in the ground and particularly rectangular and elliptical-shaped slots are used to improve the radiation directivity, gain, and efficiency. The optimized dimension of the antenna is 50W44W1.524 mm3 by using the Rogers RO4350B substrate. The measured and simulated results demonstrate that the proposed prototype has a bandwidth of 2.01 GHz (1.70–3.71 GHz) with nearly directional radiation characteristics. The highest gain of the prototype is 5.65 dBi, and the maximum 93% radiation efficiency over the frequency band. The antenna has a lower group delay, higher fidelity factor (>95%), which are significant factors for microwave head imaging. Various design structures are performed to achieve the most satisfactory result. The proposed prototype is designed and analyzed by 3D CST 2019 simulator software. Later, the antenna is fabricated and measured to examine the performance. Thereafter, a single antenna and twelve-antennas array element are configured and placed surrounding the 3D realistic shaped Hugo head model to validate the effectiveness of the single antenna as well as the designed antennas array. The antenna has satisfactory field penetration into the human head tissues with a safe SAR (specific absorption rate). Less than 0.00233 W/kg SAR is attained over the operating frequency range, which is lower than the reported antennas. The Iteratively Corrected Delay Multiply and Sum (IC-DMAS) imaging algorithm will be applied for imaging purposes. The investigation through the literature of the statistical and measured data evident that the proposed antenna is apposite for the wideband microwave head imaging applications.

An Ultrawideband (UWB) octagonal ring-shaped parasitic resonator-based patch antenna for microwave imaging applications is presented in this study, which is constructed with a diamond-shaped radiating patch, three octagonal, rectangular slotted ring-shaped parasitic resonator elements, and partial slotting ground plane. The main goals of uses of parasitic ring-shaped elements are improving antenna performance. In the prototype, various kinds of slots on the ground plane were investigated, and especially rectangular slots and irregular zigzag slots are applied to enhance bandwidth, gain, efficiency, and radiation directivity. The optimized size of the antenna is 29 × 24 × 1.5 mm³ by using the FR-4 substrate. The overall results illustrate that the antenna has a bandwidth of 8.7 GHz (2.80–11.50 GHz) for the reflection coefficient S₁₁ < −10 dB with directional radiation pattern. The maximum gain of the proposed prototype is more than 5.7 dBi, and the average efficiency over the radiating bandwidth is 75%. Different design modifications are performed to attain the most favorable outcome of the proposed antenna. However, the prototype of the proposed antenna is designed and simulated in the 3D simulator CST Microwave Studio 2018 and then effectively fabricated and measured. The investigation throughout the study of the numerical as well as experimental data explicit that the proposed antenna is appropriate for the Ultrawideband-based microwave-imaging fields.