Crowdsourcing has continued to accelerate in the biotech and life sciences sectors, with numerous biotech companies using crowdfunding to fuel their start-up ideas. Now, crowdsourcing is fast becoming an efficient means of gathering key scientific data and specimens. Just on the heels of an article last week highlighting the work of citizen scientists to gather a wide range of fungus specimens, a new report has emerged outlining a new initiative to launch an app designed to help detect melanoma, with acquired data coming from crowdsourced material.
The new NIH-funded research initiative involves both the University of Utah and Texas Tech University, with Dr. Andy J. King, assistant professor of media and communication participating from Texas Tech . Both institutions are working together on the study to help leverage what they call “mole crowdsourcing” to better diagnose possible melanomas. The ideas come from the study itself, which determined after showing 500 adult participants high-resolution images of 40 moles (nine of which were clinically diagnosed melanomas) and asking individuals to identify those they found suspicious, that the collective group had a much higher accuracy score in detecting melanomas — 90 percent versus only 50% on an individual basis.
University of Utah communication assistant professor Jakob Jensen, the senior author of the study published in the December issue of Cancer Epidemiology, conceded that “Individuals are bad at finding cancerous moles,” but added that, “Our research shows that groups of people – when looked at as a group – are very good at this task. We call this process collective effort.”
Collective effort — a hallmark of crowdsourcing — has been shown to be effective even if an individual has low reliability in performing a task, since the group effort counterbalances outliers in performance. In this case, the collective effort is not biased by itself, since the efforts are all initially made individually — it is only the total result of the individual efforts that are crowdsourced.
“Mole crowdsourcing could be implemented via modern communication technology,” added Jensen. “For instance, our research team is already working on a cell phone application that will allow people to take a photo of a mole and have that image evaluated by hundreds, if not thousands, of other users.”
While “mole crowdsourcing” is a novel concept, the researchers themselves are quick to point out that the approach — and any new “diagnostic” to determine melanomas — would never be a replacement for a traditional diagnosis from a doctor, but rather a good starting point for a person to consider seeing a dermatologist.