Applying digital technologies to the prevention, diagnosis, and treatment of stubborn conditions like cancer is a great gift of modern times. But, as we’re beginning to learn as competition heats up in this space, there are likely to be some growing pains and false starts before we achieve something that works reliably.
Mobile apps that help us catch cancers in their early stages are coming — but it’s going to take a lot of work and lots of collaboration with the medical community before they deliver consistent, accurate and actionable results.
How Are These Apps Supposed To Work?
All of this sounds like a veritable healthcare revolution — but it’s only in theory, at least for the moment. The idea is that prospective patients will eventually be able to use consumer-level technology to meet an accurate diagnosis or track changes to their bodies over time that could signify the advancement of skin and other cancers.
Researchers are focused on the following three methods to make this a reality:
- “Tele-dermatology” would let patients send photos of their skin conditions to a dermatologist for a professional opinion.
- Patients could also upload and store their photos on a secure website over a period so dermatologists can study them for changes and disease progression.
- Human dermatologists could eventually be phased out of the diagnosis process entirely thanks to fractal analysis and pattern recognition algorithms, which could spot colors and textures that signify a dangerous or cancerous condition.
Some of the results from early testing have been promising. The first example — teledermatology — was able to identify skin cancer in 88 percent of observed cases. It also proved useful in determining “benign lesions” in 97 percent of patients.
Coming in second place was fractal analysis, which identified skin cancer in a still-impressive 73 percent of cases and benign lesions in 83 percent of cases.
Why This Is Important, And Why There’s Lots Of Work Ahead
First, it’s worth noting that other forms of telemedicine are already here, making life more livable for mobility-challenged patients, veterans who don’t have access to approved hospitals and others. It’s a perfect fit in many cases, including letting psychologists and psychiatrists speak directly with their patients from across great distances. It’s a logical leap in behavioral medicine.
And it’s a potentially life-saving one, too. About 4.3 million cases of basal cell carcinoma — the most common type of skin cancer — are diagnosed in the U.S. annually, with 3000 people losing their lives to the condition.
However, “training” these apps and algorithms to correctly identify cancers and similar conditions are going to be a very long uphill climb, say researchers. That’s not to mention the technology is without potential. However, dermatologists are clear about how much work is still needed before apps like these are ready for mainstream adoption.
The first and most obvious limitation is the current lack of published clinical trials demonstrating the accuracy and usefulness of these apps. These will almost certainly come with time after the underlying technologies have been through further testing and iterations.
Medical professionals familiar with the technology are, however, very clear about the need for dermatologists to be heavily involved in app development. Although these algorithmic cancer screening tools show promise, doctors say it’s going to take much more work and a great deal of input from the medical community before they’re able to identify “fringe cases” and some of the rarer cancers.
More specifically, some melanomas and other skin conditions don’t manifest with changes in skin pigmentation. Patients relying on apps like these could receive an erroneous “all-clear” from their smartphone when, in reality, the app just isn’t equipped to detect subtler or less-common signs of the disease. Delays in treatment for any cancer could be catastrophic.
But it’s not just the lack of professional scrutiny limiting the usefulness of these apps now — it’s also the nature of the technology itself. Although photo-based pattern recognition using pocketable cameras shows some promise, it’s a poor substitute for dermatoscopes and actual tactile interactions between doctor and patient.
To put it another way, the entire intent of the app — furnishing accurate cancer diagnoses without the advice of a professional — is, for the time being, a bit of a fantasy.
What’s The Bottom Line Right Now?
When asked about the progress of apps like these, dermatologists were quick to enter the public record and advise caution. Maria Charalambides of Birmingham College said:
“Members of the public should be cautious when using such apps, as they come with risk. Any software that claims to offer a diagnostic element must be subject to rigorous testing and ongoing monitoring.”
Indeed, there are generally six phases to clinical trials, some of which take years on their own. The last stage doesn’t even take place until after the medication or application has received commercial approval from regulatory bodies. Experts like Charalambides are quick to admit the future holds a great deal of potential for telemedicine as it applies to early cancer detection. But for the time being, there are technical, practical and legal barriers to clear before this tool proves itself capable of standing on its own.
One type of app that is unquestionably important and useful, however, are education-based ones. These can alert users to especially dangerous sun conditions in their area and give feasible advice for keeping themselves and their children safe in the sun.
For the near future, you’re urged to seek a second opinion from a professional if you have even a shred of suspicion that something’s not right with your body — no matter what your telephone tells you.