Skin-deep diagnosis

Using a combination of NIR spectroscopy and skin impedance spectroscopy, Swedish scientists have potentially come up with a quick, non-invasive technique for identifying the skin cancer known as melanoma.

Melanoma is a malignant tumour of melanocytes, which produce the dark pigment melanin, and so manifests itself as a dark spot or lump on the skin. It is the fifth most common type of cancer in the UK, affecting one in 60 people, and causes the vast majority of deaths related to skin cancer. Nevertheless, the survival rate for those with melanoma is still quite high, at over 90%, providing the melanoma is diagnosed at an early stage.

At the moment, an initial diagnosis of melanoma is usually made based on a visual examination, but this needs to be done by a trained physician and is still obviously highly subjective.

So a team of Swedish scientists led by Britta Lindholm-Sethson at Umeå University decided to see whether a combination of NIR spectroscopy and skin impedance spectroscopy, which monitors the flow of electrical current between two electrodes placed on the skin, could provide a more objective form of diagnosis. Using these two spectroscopy techniques, they probed the skin lesions on 50 patients, with the aim of distinguishing malignant melanomas from harmless moles.

After combining the spectral data from the two techniques, Lindholm-Sethson and her team analysed the data using principal component analysis and partial least squares discriminant analysis. As they report in a paper in Skin Research and Technology, they found they could distinguish between the 12 malignant melanomas and 38 moles with an accuracy of up to 95%.

Interestingly, they were also able to distinguish between different forms of melanoma and different types of mole; the spectral data even indicated that a type of mole known as a dysplastic naevi might actually be an early stage of skin cancer.

Blog tags: