Mole Tracker - The pitch

Tracking the health of moles is too stressful and too expensive. With mole tracker it’s a simple quick experience with instant results. It’s saves doctor’s time and patients’ stress.

Mole Tracker - The fun part

The hardware design would be fun, but I’m more interested in the classification problem. Given the features of a mole, what is the probability that it is malignant? What’s nice is that it could train on data from biopsies and historical data about moles not needing to be biopsied within the following few years.

Mole Tracker - The growth potential

There are around 10,000 dermatologists in the US. If one of these devices is needed for every 5 dermatologists and needs replacing every 5 years, that amounts to $800,000 a year in revenue per year. If this went global, the revenue could be around $10 million a year.

Mole Tracker - The monetization

The hardware and the software to interpret the results would be sold as a package and be designed to integrate well with existing medical care software. The system could reduce the need to have as many dermatologists (or let them see more patients). I think this could reduce the amount of time it takes to evaluate a patient by about 10 minutes and only take 1 minute. In a dermatology department with 5 doctors this could save 4 hours of doctor time a day. Over a year that adds up to $50,000 (assuming $50 an hour for a doctor’s time). I think the device would sell for only $50,000 and last around 5 years. The amount of savings is $200,000 over the life of the device.

Mole Tracker - The idea

Instead of having to see a dermatologist, it would make more sense to have a patient’s body scanned for moles, the location of the moles noted, and detailed pictures taken. This could be done by having a Microsoft Kinect sensor and a camera mounted on a motorized arm that moves around the patient's body, while they hold an awkward TSA-like stance. The pictures of the moles would be stored with relative body coordinates so they could be overlayed on a digital manikin (so no sensitive area photos are stored).

The moles then could be analyzed for features that are known to be correlated with melanoma. One of the most important things is that any year-over-year change can easily be tracked with these images, which is important because change/growth is the best indicator of a problem. Another thing this system could do is advise which moles should be watched more carefully in the future and perhaps even give an estimate for how long before another scan may be needed.

Mole Tracker - The motivation

Skin cancer is a scary thing. For many fair-skinned people, doctors recommend having their moles checked for concerning signs yearly or every few years. One of the signs that something is wrong is changes in a mole. Unfortunately, most people don’t remember exactly what their moles looked like a few years ago.