Synthetic Ultra-High-Resolution Millimeter-Wave Imaging for Skin Cancer Detection

IEEE Trans Biomed Eng. 2019 Jan;66(1):61-71. doi: 10.1109/TBME.2018.2837102. Epub 2018 May 17.

Abstract

This work introduces, for the first time, a millimeter-wave imaging system with a "synthetic" ultra-wide imaging bandwidth of 98 GHz to provide the ultra-high resolutions required for early-stage skin cancer detection. The proposed approach consists of splitting the required ultra-wide imaging bandwidth into four sub-bands, and assigning each sub-band to a separate imaging element, i.e., an antenna radiator. Each of the sub-band antennas transmits and receives signals only at its corresponding sub-band. The captured signals are then combined and processed to form the image of the target. For each sub-band, a Vivaldi tapered slot antenna fed with a combination of substrate-integrated waveguide and coplanar waveguide is designed and microfabricated. Design techniques are also provided for the four similarly-shaped sub-band antennas for achieving excellent impedance matches ( S11 < -10 dB) and nearly constant gains of 10 dBi over the entire 12-110 GHz bandwidth. The design procedure is validated by comparing the simulated results with measurements performed on the fabricated prototypes. Excellent agreements are obtained between simulations and measurements. Finally, the feasibility of detecting early-stage skin tumors in three dimensions is experimentally verified by employing the sub-band antennas in a synthetic ultra-wideband imaging system with a bandwidth of 98 GHz. Two separate setups, each comprising a dispersive skin-mimicking phantom as well as two dispersive spherical tumors, are constructed for imaging experiments. Lateral and axial resolutions of 200 μm are confirmed, and a successful reconstruction of the spherical tumors is achieved in both cases.

Publication types

  • Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

  • Algorithms
  • Diagnostic Imaging / instrumentation
  • Diagnostic Imaging / methods*
  • Equipment Design
  • Humans
  • Image Interpretation, Computer-Assisted / methods
  • Phantoms, Imaging
  • Skin / diagnostic imaging*
  • Skin Neoplasms / diagnostic imaging*