“Keep an eye on that one, and call me if it changes,” is as much a stock phrase for dermatologists as “Wear your sunscreen.” Yet, how do patients know if a mole changes? I’m quite sure many of my patients wouldn’t notice if I shaved my head and grew a beard, let alone notice if 1 of 30 moles on their back changed color.
Mole Mapper is an iPhone app developed by the department of dermatology at Oregon Health and Science University (OHSU) to solve this problem. The app provides a framework for patients to photo, measure, and track their moles. With clear instructions, an anatomical map, and sophisticated markers, it is a significant aid for motivated patients who want medical-grade photos suitable for tracking.
To standardize the photos, the app prompts you to include a nickel, dime, or quarter in photos with nevi of interest. The user then calibrates the app by pinching onscreen circles overlying the photo such that they correspond exactly to the circumference of the coin and to the mole. Using a coin as a standard, the app then calculates the precise size of the mole regardless of the size of the photo. For example, photos taken 2 feet and 4 feet away both give the same diameter because both photos are calibrated by the dime in each.
The app was developed by a cancer biologist, Dan Webster, Ph.D., to help his wife monitor her moles between dermatology appointments. Interestingly, it was largely developed by a single person, a sign that creating apps is nearly entering into a DIY era. This increases the possibility for useful health care tools to be developed while also increasing the already crushing crowd of apps, few of which are truly useful.
The app’s functionality would not have been possible without the inclusion of Apple’s ResearchKit and Sage Bionetworks’ Bridge Server. ResearchKit provided open-source tools to facilitate informed consent over the phone and the ability to conduct participant surveys, among other activities. Bridge Server enabled the app to encrypt and securely transfer participant data from the phone to firewalled storage. The combination of these two software frameworks is paving the way for an exciting future of integrated technology and biomedical research.
According to Dan Webster, “ResearchKit is a game-changer because it provides an open-source platform for elegant informed consent, measurement tools, and participant data protection. The ability for participants to have so few barriers to contribute to a research study is the truly transformative aspect of ResearchKit, and we have seen unprecedented numbers of research study enrollees as a result.”
But that’s not all. The app is more than just a consumer tool for tracking – ResearchKit allows OHSU researchers to gather data on nevi, track them over time, and learn characteristics associated with melanoma from user-generated outcomes. This could significantly increase our understanding of melanoma and perhaps spawn an artificially intelligent app that learns to diagnose melanoma without human assistance.
Because of the institutional review board’s requirements for their research, users must be 18 years old to participate in the study. The app gracefully walks users through the consent process and even has a knowledge check at the end to ensure that they understood the risks and benefits of participating. The consent process is so streamlined that it ought to be a model for us to consent any patient for any reason.
To be clear, the app does not make diagnoses. It only provides a framework for patients to photograph their moles and track them. It also politely prompts users to rephotograph moles every 30 days so changes can be recorded.
There are apps with similar names, so be sure you have Mole Mapper from Sage Bionetworks. I tried it out to offer my experience here. Taking photos was as simple as any photo on an iPhone. Like any selfie, however, there are azimuth limits to the human arm – you can’t get shots in remote bodily corners easily. Also, placing a coin on yourself is easier said than done, unless you want to use your bubble gum to hold it in place while you take the shot. (I asked for assistance from my wife instead.)
The photos I took were accurate when compared with the measured diameter in real life, but there are still user-dependent adjustments that could lead to large artifacts. Making the measurement circles even slightly smaller or larger around the coin or the mole can lead to more than a millimeter of margins of error. If detecting melanoma requires less than 1-mm error in mole changes, then this could limit its usefulness.