By combining imagining with digital analysis and machine learning, researchers have developed by an automated technology to help physicians detect melanoma at its early stages. Melanoma can be hard to detect and diagnose. As the skin cancer can start by being a mere mole which later turns out to be a cancerous tumour.
James Krueger, Professor at Rockefeller University in the United States stated, “There is a real need for standardization across the field of dermatology in how melanomas are evaluated.”
Krueger also added, “Detection through screening saves lives but is very challenging visually, and even when a suspicious lesion is extracted and biopsied, it is confirmed to be melanoma in only about 10% of cases.”
The new approach involves computer algorithms that process the images of lesions and extract quantitative data from them like shape and colour of lesions. Then, they generate an overall risk score that is called a Q-score, indicating the chances of developing melanoma skin cancer. A score between zero and one represents the higher probability of lesion being a cancerous tumour.
Researchers claim that the new method can detect early melanomas on skin with 98% accuracy, which is impossible to achieve with conventional methods.
“The success of the Q-score in predicting melanoma is a marked improvement over competing technologies,” said Daniel Gareau, instructor in clinical investigation in the Krueger laboratory.
As previous studies have shown, the number of colours in a lesion turned out to be the most significant biomarker for determining malignancy.
Some biomarkers were significant only if looked at in specific colour channels – a finding the researchers say could potentially be exploited to identify additional biomarkers and further improve accuracy.
“I think this technology could help detect the disease earlier, which could save lives, and avoid unnecessary biopsies too,” said Gareau.
The research was published in the journal Experimental Dermatology.