Microwave breast cancer detection via cost-sensitive ensemble classifiers: Phantom and patient investigation

2017 ◽  
Vol 31 ◽  
pp. 366-376 ◽  
Author(s):  
Yunpeng Li ◽  
Emily Porter ◽  
Adam Santorelli ◽  
Milica Popović ◽  
Mark Coates
Keyword(s):  
Breast Cancer ◽  
Cancer Detection ◽  
Breast Cancer Detection ◽  
Ensemble Classifiers ◽  
Microwave Breast Cancer Detection

scholarly journals The Performance Comparison of a Dual-Ridge Horn Antenna and a Planar Monopole Antenna in the Microwave Breast Cancer Detection

2020 ◽  
Vol 9 (2) ◽  
pp. 84-92
Author(s):  
A. R. Celik ◽  
M. B. Kurt
Keyword(s):  
Breast Cancer ◽  
Dielectric Constant ◽  
Cancer Detection ◽  
Breast Cancer Detection ◽  
Monopole Antenna ◽  
Ultra Wideband ◽  
Microwave Measurement ◽  
Significant Difference ◽  
Planar Monopole Antenna ◽  
Microwave Breast Cancer Detection

Detection of the breast cancer tumors at an early stage is very crucial to be successful in the treatment. Microwave measurement systems have gained much attention for this aim over last decades. The main principle of these systems is based on the significant difference in the dielectric properties of the malignant tumor and normal breast tissue in the microwave frequencies. In this paper, firstly several breast cancer detection techniques are mentioned. Then the advantages of the using microwaves in the detection systems are given. After that, some simulation and experimental studies of the radar-based ultra-wideband microwave measurement system are presented to detect tumor. The main purposes of these measurements are comparing the performance of a previously designed planar monopole antenna (PMA) with a dual-ridge horn (DRH) antenna and demonstrating a simple microwave breast cancer detection system. In the system, a planar breast phantom which is consisted of low dielectric constant material to represent the healthy tissue and high dielectric constant material to represent the tumor is used. Firstly, the measurements are made without tumor in the phantom. Then, the tumor-mimicking object is located to the phantom. In the measurements, both the PMA and DRH antennas are used respectively. These antennas are ultra-wideband and directional. They have narrow beamwidth and stable directional pattern at the interval of 3-10 GHz. According to the return loss results, the reflected energy increases when the antenna gets close to the tumor. Therefore, it can be said that the scattering parameters give important information about the tumor. According to the obtained results in this study, it can be said that the performance of the compact-sized PMA is better than the DRH antenna having larger size.

Breast cancer imaging and detection using microwave technique

2020 ◽  
Author(s):  
◽  
Ahmed Maher Abed
Keyword(s):  
Breast Cancer ◽  
Dielectric Properties ◽  
Cancer Detection ◽  
Wide Band ◽  
Breast Cancer Detection ◽  
Ultra Wide Band ◽  
Late Time ◽  
Simulation Tool ◽  
Esprit Algorithm ◽  
Microwave Breast Cancer Detection

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI-COLUMBIA AT REQUEST OF AUTHOR.] Worldwide, breast cancer continues to be the top cause of death among women and the second-leading cause of cancer death after lung cancer. Thus, it has become a great global concern. Years of research on both diagnostic and therapeutic breast cancer detection and imaging using microwave techniques has resulted in a variety of novel approaches and studies. These approaches and studies utilize numerical breast phantoms that model structural complexities, tissue heterogeneity, and dispersive dielectric properties. In this dissertation, a microwave breast cancer detection technique was investigated and Ultra-Wide Band (UWB) radar imaging was used. A UWB antenna was designed and modeled using CST Microwave Studio and was used for Ultra-Wide Band microwave breast cancer detection. A new calibration approach for microwave breast cancer detection was proposed to calibrate the signals before applying the beamforming algorithms. A simulation was also used to validate the proposed techniques. Two signal calibration approaches were proposed to remove the high magnitude clutter from the signals. The two approaches are based on the state-space method Autoregressive Moving Average (ARMA). The first approach is derived from Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT). This algorithm is referred to as the Pole Isolation via ESPRIT algorithm (PI-ESPRIT). The second approach is also derived from a previously proposed algorithm for microwave breast cancer detection. Th second calibration algorithm is referred to as the Modified Pole Removal algorithm. The Modified Pole Removal algorithm works to detect tumors, in contrast to the previously proposed algorithm (Pole Removal) that shows a lack of tumor detection. Three beamforming techniques were used to focus the signals onto the voxels through the breast phantom. Another beamforming algorithm was proposed, along with the Transmitting-Receiving Antenna Separation Distance (TRASD), which allows for the reduction of the late time clutter effect and improvement of the Signal to Clutter Ratio (SCR) when using the PI-ESPRIT algorithm. Using CST simulation tool, antennas arrayed around the breast are designed to simulate the transmitting/receiving signals. Numerical phantoms, with complicated structures and dispersive dielectric properties of breast tissue, have been developed with the CST simulation tool for simulating electromagnetic propagation. The received signals are imported into the MATLAB program to investigate the proposed approaches and compare them to conventional approaches. Overall, the results have shown improvement.

Sign in / Sign up

Export Citation Format

Share Document