Crowdsourcing could help to ID suspicious moles

Researchers may have found a new technique for identifying suspicious-looking moles: crowdsourcing.

Study investigators from the University of Utah, Salt Lake City, and Texas Tech University, Lubbock, Texas, say they have found a new technique for identifying suspicious-looking skin moles: crowdsourcing. Crowdsourcing refers to using crowds of people, often recruited online, to accomplish tasks or solicit services.

For their study, the researchers showed 500 adults high-resolution images of 40 moles - nine of which were clinically diagnosed melanomas - and asked them to circle those that looked suspicious. Although the average person was only able to identify about half of the melanomas, researchers found that when looking at the group collectively, 19 percent of the participants were able to correctly identify 90 percent of the melanomas.

This process, which the researchers call “collective effort,” can be effective even when a single person has low reliability at a task, as the pattern of the group overcomes the limitations of each person. Collective effort does not utilize group decision-making, as respondents independently perform each task.

Senior author Jakob Jensen, Ph.D., a University of Utah assistant professor of communication, and colleagues refer to the collective-effort approach as “mole crowdsourcing.”

“One of the nice things about this finding is that it suggests several exciting avenues for practice and research,” Dr. Jensen tells Dermatology Times. “From a practice standpoint, it highlights the potential value of encouraging patients to talk with others following skin self-examination. From a research standpoint, our data suggests that mole crowdsourcing may be considerably more effective than traditional skin self-examination. With a 19 percent cut-off, our ‘crowds’ managed to find 90 percent of melanomas.”

The researchers stress that mole crowdsourcing would not replace the importance of clinical skin examinations, but provides an evidence-based strategy for funneling high-risk patients toward those resources.

The findings were published in the December issue of Cancer Epidemiology.