Detecting skin malignancy using elastic light scattering spectroscopy


CANPOLAT M., Akman A., Ciftcioglu M. A., ALPSOY E.

Conference on Diagnostic Optical Spectroscopy in Biomedicine IV, Munich, Germany, 19 - 21 June 2007, vol.6628 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 6628
  • Doi Number: 10.1117/12.729463
  • City: Munich
  • Country: Germany
  • Keywords: skin, spectroscopy, non-invasive, cancer, detection, OPTICAL COHERENCE TOMOGRAPHY, IN-VIVO, DIAGNOSIS
  • Akdeniz University Affiliated: Yes

Abstract

We have used elastic light scattering spectroscopy to differentiate between malign and benign skin lesions. The system consists of a UV spectrometer, a single optical fiber probe and a laptop. The single optical fiber probe was used for both delivery and detection of white light to tissue and from the tissue. The single optical fiber probe received singly scattered photons rather than diffused photons in tissue. Therefore, the spectra are correlated with morphological differences of the cells. It has been shown that spectra of malign skin lesions are different than spectra of benign skin lesions. While slopes of the spectra taken on benign lesions or normal skin tissues were positive, slopes of the spectra taken on malign skin lesions tissues were negative. In vivo experiments were conducted on 20 lesions from 18 patients (11 men with mean age of 68 +/- 9 years and 7 women with mean age of 52 +/- 20 years) applied to the Department of Dermatology and Venerology. Before the biopsy, spectra were taken on the lesion and adjacent (approximately 1 cm distant) normal-appearing skin. Spectra of the normal skin were used as a control group. The spectra were correlated to the pathology results with sensitivity and specificity of 82% and 89%, respectively. Due to small diameter of fiber probe and limited number of sampling (15), some positive cases are missed, which is lowered the sensitivity of the system. The results are promising and could suggest that the system may be able to detect malignant skin lesion non-invasively and in real time.