An Early Breast Cancer Detection by Using Wearable Flexible Sensors and Artificial Intelligent

Abstract

Wearable devices are currently of great importance in developing health management and care applications and devices especially for early breast cancer detection (BCD). This is supported by the novel innovations in materials and techniques to construct highly accurate and comfortable biosensors for wearables. In this paper, the proposed flexible sensor is integrated with a Bra to realize a wearable breast cancer detection system. The proposed antenna-based sensor is composed of a CPW monopole antenna with an overall compact size of $24\times45$ mm2 on flexible PCB Roger substrate with thickness 0.17 mm. The proposed sensors have enough bandwidth from 1.5 to 8 GHz at -6 dB reflection coefficient and conformal structure for biological structures and biocompatibility. The specific absorption rate (SAR) has been calculated and measured for the proposed sensor with a value of 0.75 W/kg at 0 dBm to ensure safety level. For testing, real shaped rubber phantoms from medical school enables the dynamic combination of breast and tumor to create test scenarios for breast cancer detection. $2\times 2$ antenna-based sensors elements are placed around the breast phantom to gather data on scattering parameters for tumor characterization. Several simulation and measurement scenarios are presented to validate detection, optimum number of sensors to be used as well as training data for developed detection algorithms. Artificial intelligence techniques are used among which the CAT-Boost technique for scanning collected data and identifying the undesired tumor component inside the breast.

Keywords

• Coplanar waveguide (CPW) wearable
• flexible sensors
• breast cancer detection (BCD)
• machine-learning algorithms (MLA)
• specific absorption rate (SAR)
• conformal