A hand-held polarized-light camera for the detection of skin cancer borders

Jessica C. Ramella-Roman, Ken Lee, Steven L. Jacques

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Scopus citations


Polarized light imaging can be used to map the borders of skin cancer that are invisible to the human eye. We designed a hand-held polarized-light camera that is sensitive to the superficial layers of the skin where cancer often arises. The camera system consists of two 8-bit CCD mounted on two sides of a polarizing beam splitter. An imaging lens mounted on the beam splitter collects light reflected from the skin. The collected light is divided into two states of polarization: parallel (PAR) and perpendicular (PER) to the incident light orientation. A custom-code combines images streaming from both cameras to yield an image based on the polarization ratio (PAR - PER)/(PAR + PER), which is sensitive to the superficial skin layer. Experiments in reflectance mode on micro-spheres solutions were conducted to test the system; Monte Carlo simulations of these experiments demonstrated excellent correlation. Early clinical work was conducted at the Oregon Health and Science University in the Dermatology department. Images of several skin lesions are presented.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
EditorsL.M. Hanssen
Number of pages7
StatePublished - 2003
Externally publishedYes
EventOptical Diagnostic Methods for Inorganic Materials III - San Diego, CA, United States
Duration: Aug 6 2003Aug 7 2003


OtherOptical Diagnostic Methods for Inorganic Materials III
Country/TerritoryUnited States
CitySan Diego, CA


  • Polarized light
  • Skin cancer

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics


Dive into the research topics of 'A hand-held polarized-light camera for the detection of skin cancer borders'. Together they form a unique fingerprint.

Cite this