June 27, 2017 | Written by: Michelle Cullen
Categorized: AI/Watson | IBM: The Backstory
Share this post:
Summer began this past week in the Northern Hemisphere. As people get ready for more outdoor activities and the increased sun exposure that comes with the season, it’s important to remember to remain sun-smart.
Around the world, there are over a million new cases of skin cancer diagnosed each year. While melanoma is the least common form of skin cancer, accounting for just one percent of diagnoses, it is responsible for the majority of skin cancer deaths.
Since 1930, there has been a 2000 percent increase in melanoma. Globally, one person dies each hour from melanoma, and the incidence is rising.
In the United States, the incidence of melanoma has doubled since 1973, with a similar increase in the United Kingdom. The highest incidence of melanoma can be found in Australia and New Zealand, where it is more than twice that of North America.
Combatting melanoma, the deadliest form of skin cancer
To address this rising threat, Cancer Council Victoria created the “Slip, Slop, Slap!” campaign in the early 1980’s, one of the most successful health campaigns in Australia’s history. This message was developed to encourage Australians to “slip on a shirt, slop on sunscreen and slap on a hat.” It has since been expanded to include “seek shade and slide on sunglasses”.
Technology is also being applied in this battle. IBM, in partnership with the skin cancer detection program MoleMap and the Melanoma Institute of Australia, is researching how IBM Watson may potentially help clinicians with early detection. Together in a research project to test efficacy, the organizations are working to train IBM Watson on how to recognize cancerous skin lesions so that clinicians can make a more accurate diagnosis.
Early detection of melanoma increases a patient’s chance of survival, yet this isn’t an easy task for doctors. In general, a clinician is accurate about 60 percent of the time. If a high-powered digital microscope is used by a trained clinician, diagnosis accuracy can increase to more than 80 percent.
With IBM Watson, the research team is hoping to achieve 90 percent accuracy of identification of cancerous skin lesions. Ideally, they would like to be able to train IBM Watson to potentially help clinicians better identify if a lesion is melanoma, another form of skin cancer, or a different skin disease altogether.
When early identification results in prompt treatment, it could potentially contribute to saving lives and hundreds of millions of Australian federal government dollars.
Experiencing the Australian sun first hand
Having lived in Australia for several years, I witnessed the seriousness with which Australians take precautions against skin cancer. I remember going to the beach and watching them layer on SPF, hats and shirts. Whereas in my home country, the United States, sun bathing for a tan is still pervasive.
My experience with the Australian sun was relatively brief. However, Andrea Acton, a fellow IBMer, has a very personal story to share about how this disease has affected her life.
You can also learn more about IBM Watson and how it is helping clinicians improve skin cancer diagnosis in this video.
IBM’s statements regarding its plans, directions and intent are subject to change or withdrawal without notice at IBM’s sole discretion.