4 ms·
[I've written about this on HN a couple times before, but it was a long time ago and I can't find it in my history.] I designed a software program to detect sk
by Shooter 17y ago
[I've written about this on HN a couple times before, but it was a long time ago and I can't find it in my history.]
I designed a software program to detect skin cancer a few years ago. At the time, there were probably already a dozen such skin cancer pattern recognition programs/systems for sale. Almost all of them are/were significantly more accurate at detecting melanomas than dermatologists. I was mainly just trying to improve the accuracy of the existing software a tiny bit and to improve my programming skills. I didn't work on the hardware side of the equation at all. I didn't expect to make any money or to achieve much...I'm not too sharp as far as algorithms go and I usually coast by on marketing and domain knowledge. But, miraculously, I did eventually achieve my accuracy/consistency goals. I was unbelievably excited. I formed a small company to handle the IP for all my medical software experiments - including the skin cancer software - and I set about marketing the software to medical companies with dollar signs in my eyes. And it was one of the few times I DIDN'T run into crazy patent issues. I eventually sold the company to one of the largest engineering conglomerates in the world (to their medical division, specifically) so that I wouldn't have to deal with the regulatory crap. Being greedy, I worked out a royalty scheme on the skin cancer software in exchange for a smaller company sale price.
SERENDIPITOUS SIDE STORY: I was working on the software one day while my wife was getting ready to go to an amusement park, and I noticed she had a mole that reminded me of one of my cancer training slides. I asked her to get it biopsied. She put it off for a week or two, but I kept nagging her that I didn't like the looks of it. She went to her GP, and they told her it was fine and not to worry about it. I was pissed off, and asked her to go to a dermatologist to have it looked at. She went to a dermatologist...and was again told that it was fine and not to worry about it. I demanded she have it biopsied...to ease my mind. My wife works at the same hospital with these physicians, and she didn't want to rock the boat...but I'm very annoying when I get stuck on something. So, my wife finally had it biopsied. It comes back as a particularly nasty melanoma. She ends up having several large excisions and her lymph nodes biopsied. They eventually completely removed her axillary lymph nodes on one side (which caused some other problems), but she was extremely lucky overall. She had PET scans and other monitoring for a bit. Then they said she is completely cancer-free. They do exams every other month for her now, because she is pregnant - and pregnancy increases melanin production. Melanoma also has a large genetic component, which I didn't really know about until all of this happened. Her uncle died of the same type of melanoma yesterday. And they had told him HE was completely cancer-free about 10 months ago. This shit can keep you up at night.
My plans to move to a tropical paradise have been derailed, unfortunately, but my wife is still cancer-free.
So...in the few years since the sale I've made exactly $0 in skin cancer software royalties. It turns out the huge conglomerate bought my company for the insurance/billing stuff we had designed, not the skin cancer software. They have been largely unsuccessful in selling complete skin cancer detection systems that they already own, and don't intend to do much in that area for the foreseeable future. Most dermatologists simply don't want the software/devices, for starters.
- DaniFong 17y agoWow, Shooter, what a story. Congratulations on writing that software, achieving your accuracy/consistency goals, and selling your company, though I share your disappointment with the fact that it never came out as a real product. It's probably difficult to sell to dermatologists something that sees what they see and does it better. That's why we were interested in the much more challenging technical environment of using software on consumer devices (like handheld cameras and cellphones). If you have the time, I'd really like to chat some more about your experiences and your work. My friend, Adam Stein, who once worked for the FDA that I mentioned is looking for a cofounder in this area, and says that he's learned the ropes to deal with the regulatory issues. As for the side story involving your wife, again, wow. I hear too many stories like this. Medical experts simply don't have a big enough stake in your health to be counted upon. Health is something you have to take in your own hands. I am so glad that you did.