Hyperspectral camera shows promising results in detection of skin cancer
- Date:
- February 27, 2014
- Source:
- Technical Research Centre of Finland (VTT)
- Summary:
- A lightweight, handheld, ultra-precision hyperspectral camera has been developed for the detection of skin cancers and their precursors. From the surface of the skin, the camera recognizes early stages of cancer that are invisible to the naked eye. The hand-held, mobile hyperspectral camera images the skin region in two seconds. The large field of view (12 cm2) enables the detection of large skin areas at once. The preliminary results are promising.
- Share:
VTT Technical Research Centre of Finland has developed a lightweight, handheld, ultra-precision hyperspectral camera for the detection of skin cancers and their precursors. From the surface of the skin, the camera recognises early stages of cancer that are invisible to the naked eye. Collaborators in the pilot study are the University of Jyväskylä, the Päijät-Häme Central Hospital and the Skin and Allergy Hospital of Helsinki University Central Hospital. The preliminary results are promising.
The hand-held, mobile hyperspectral camera images the skin region in two seconds. The large field of view (12 cm2) enables the detection of large skin areas at once.
In the pilot study, the camera has been used to detect the skin areas with field cancerization i.e. areas of multiple skin cancer precursors, actinic keratoses, for early treatment of the affected areas. The hyperspectral camera has also been used to detect the borders of poorly delineated skin tumours, such as lentigo malignas, which are difficult to detect by the naked eye, in order to avoid the need for re-excisions. The preliminary results are promising.
Developed by VTT on the basis of the Fabry-Perot interferometer, the hyperspectral camera captures images in up to 70 narrow wavelengths, whereas a regular camera uses only three. The spectral image generated is a three-dimensional cube built of numerous layers of greyscale images, each of which has been taken within a limited wavelength range. A spectrum for each pixel of the spectral image is formed by the images within the cube. Different biological tissues can be identified by their reflected spectra in hyperspectral images. Computational methods are used to interpret these images, in order to determine the position and size of the tumour to be treated. In the ongoing pilot study, all results are being verified by histopathological sampling.
Story Source:
Materials provided by Technical Research Centre of Finland (VTT). Note: Content may be edited for style and length.
Cite This Page: