AI-assisted 3D mobile teledermoscopy, warfare technology and genomics are being deployed in the fight against melanoma, driving early detection and lowering death rates, according to the Australasian College of Dermatologists.
About 1770 people die from melanoma every year in Australia, but this week the leading international melanoma researchers and clinicians have gathered in Brisbane for the 9th World Congress of Melanoma in Brisbane to showcase new medical and scientific discoveries revolutionising prevention, diagnosis and treatment.
“To improve melanoma outcomes, the focus on prevention, early detection and new treatment strategies must continue,” Professor H. Peter Soyer, dermatologist with the Australasian College of Dermatologists, said.
“Precision medicine for melanoma utilising the novel technologies of genome sequencing combined with well-known phenotypic characteristics, 3D teledermatology and mobile teledermoscopy assisted by artificial intelligence will allow us to develop protocol-driven support systems.”
Most invasive melanomas in Australia are diagnosed at an early stage (less than 1mm thick) with a 96 per cent overall survival of 20 years. Early detection is key and imaging technology and artificial intelligence in the traditional clinical setting, and via mobile teledermoscopy and melanoma apps will continue to improve survival rates.
In a recent study, 87 per cent of participants said mobile teledermoscopy — which allows patients to take images of skin lesions to send to dermatologists for remote diagnosis — improved how well they examined their skin, 94 per cent said the dermatoscope was easy to use and 86 per cent were motivated to examine their skin more often.
In an article published Monday in the Medical Journal of Australia, Professor Soyer and QUT’s Professor Monika Janda wrote that applying high technology solutions to the difficult task of selecting and monitoring moles was improving survival.
“Adopting military surveillance and warfare technology, there are computer algorithms that search for changes in moles’ appearance over time. Deep convolutional neural networks analysis can group them into benign or malignant lesions with high accuracy,” Janda and Soyer wrote.
In one study, they wrote, an algorithm had a better sensitivity and specificity performance compared with the average of 21 dermatologists in identifying melanocytic lesions.
But Professor Sayer said more work needs to be done to eradicate melanoma.
“Research into more personalised, technology driven care for identified individuals who have a higher risk of developing melanoma is needed. The focus on early detection must continue through public health campaigns, consumer engagement and medical education,” he said.