Abstract

Terahertz (THz) imaging technology has shown great application potential in biological testing due to its low photon energy and broadband spectral features. After integrating a metamaterial sensor, the THz field-biological tissue interaction is dramatically enhanced, so that image contrast is significantly improved.

In this work, THz focal-plane biological imaging is implemented based on a metamaterial sensor composed of a slot array. By adopting this technique, THz spectral images of several plant leaves and animal tissues are obtained, achieving an approximate 3-fold enhancement in image contrast compared with those acquired using a bare silicon wafer.

Simultaneously, the focal-plane detection mode greatly reduces time consumption compared with the traditional raster-scanning mode. In addition, a deep learning algorithm is applied as a spectral classifier, which effectively segments different tissue regions and further improves image quality. This work demonstrates the application value of the technique as a biological testing platform.