Correcting Cherenkov light attenuation in tissue using spatial frequency domain imaging for quantitative surface dosimetry during whole breast radiation therapy

J Biomed Opt. 2018 Nov;24(7):1-10. doi: 10.1117/1.JBO.24.7.071609.

Abstract

Imaging Cherenkov emission during radiotherapy permits real-time visualization of external beam delivery on superficial tissue. This signal is linear with absorbed dose in homogeneous media, indicating potential for quantitative dosimetry. In humans, the inherent heterogeneity of tissue optical properties (primarily from blood and skin pigment) distorts the linearity between detected Cherenkov signal and absorbed dose. We examine the potential to correct for superficial vasculature using spatial frequency domain imaging (SFDI) to map tissue optical properties for large fields of view. In phantoms, applying intensity corrections to simulate blood vessels improves Cherenkov image (CI) negative contrast by 24% for a vessel 1.9-mm-in diameter. In human trials, SFDI and CI are acquired for women undergoing whole breast radiotherapy. Applied corrections reduce heterogeneity due to vasculature within the sampling limits of the SFDI from a 22% difference as compared to the treatment plan, down to 6% in one region and from 14% down to 4% in another region. The optimal use for this combined imaging system approach is to correct for small heterogeneities such as superficial blood vessels or for interpatient variations in blood/melanin content such that the corrected CI more closely represents the surface dose delivered.

Keywords: Cherenkov; imaging; radiation therapy; spatial frequency domain imaging; tissue optics.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Breast / diagnostic imaging
  • Breast Neoplasms* / diagnostic imaging
  • Breast Neoplasms* / radiotherapy
  • Equipment Design
  • Female
  • Humans
  • Image Interpretation, Computer-Assisted / methods*
  • Phantoms, Imaging
  • Radiometry / instrumentation
  • Radiometry / methods*
  • Signal Processing, Computer-Assisted*