[Description of differential optical spectropolarimetric imaging system (DOSI) for non-invasive detection of skin tumors].

BACKGROUND

Most physicians diagnose skin malignancy using theoretical criteria and clinical experience. Apart from dermatoscopy, noninvasive means for skin tumors' diagnosis are highly expensive and are not in daily use. We developed a simple, handy and relatively inexpensive tool for non-invasive diagnosis of skin tumors. The Differential Optical Spectropolarimetric Imaging system (DOSI) is based on collecting information from the surface and depth of skin tumors, using the advantage of liquid crystal devices (LCDs].

OBJECTIVE

DOSI feasibility study in order to diagnose skin tumors non-invasively.

METHOD

Skin tumor optical data was collected in order to create image indicators for specific skin tumors. Spectral images of skin tumors are captured before the lesion that was scheduled for removal was surgically excised. Repeating features of images of each tumor type (which was histologically diagnosed) that characterize the specific tumor and show distinction from the features of other tumors are recorded.

RESULTS

The feasibility of DOSI has been demonstrated for research, but it is not yet mature for clinical use. Specific images, indicating specific skin tumors could not yet be reliably concluded. One relatively obvious characteristic for malignancy seems to be crescent-like and finger-like growth patterns. In the non-malignant lesions: crescent and finger-like structures are of less pronounced orders of magnitude. Crescents are more open and radius of curvature are of larger orders of magnitude than in malignant specimen.

DISCUSSION

The advantage of using LCD is the ability to manipulate light electronically, without any movable components. LCDs are the building blocks of optical filters and polarization controllers. Using optical filters and polarization controllers, we are able to electronically manipulate the polarization and wavelength of our light source without any movable components. The light source emits different wavelengths (colors) that penetrate the tissue at different depths. The camera (CCD) captures the diffused reflected light from a skin lesion as a sequence of images for each polarization stage. By applying our novel method for each polarization stage we are able to overcome the noise of the diffused reflection and emphasize the scattering areas. Emphasizing the scattering areas provided images that may indicate malignancy.

CONCLUSIONS

The feasibility of DOSI has been demonstrated. In order to obtain reliable diagnostic image indicators for specific skin tumors, more clinical tests are necessary.