Scientists Create Algorithm That Paves Way For Skin Cancer Diagnoses Via Smartphone
- 7 years ago
Smartphones can perform all sorts of wonderful tasks, yet, computer scientists from Stanford University envision a future where the devices can do something truly astounding and possibly life-changing.
Smartphones can perform all sorts of wonderful tasks, yet, computer scientists from Stanford University envision a future where the devices can do something truly astounding and possibly life-changing.
They’re working on turning phones into mobile skin-cancer-diagnosis centers.
The group is already rather far along, having put together an algorithm with an identification rate that is on par with board-certified dermatologists.
Getting to that point required a great deal of work, including the gathering and sorting of tens of thousands of images showing the various types and severities of the skin affliction.
Brett Kuprel, one of the team members, noted, “There’s no huge dataset of skin cancer that we can just train our algorithms on, so we had to make our own.”
Now that the skin cancer identification means has proven largely successful, the team is looking forward to adjusting it in a way that makes it compatible with smartphones.
The scientists hope that they will be able to do so in the near future, as skin cancer presents a significant health risk, with 5.4 million new cases in the U.S. being reported every year.
Smartphones can perform all sorts of wonderful tasks, yet, computer scientists from Stanford University envision a future where the devices can do something truly astounding and possibly life-changing.
They’re working on turning phones into mobile skin-cancer-diagnosis centers.
The group is already rather far along, having put together an algorithm with an identification rate that is on par with board-certified dermatologists.
Getting to that point required a great deal of work, including the gathering and sorting of tens of thousands of images showing the various types and severities of the skin affliction.
Brett Kuprel, one of the team members, noted, “There’s no huge dataset of skin cancer that we can just train our algorithms on, so we had to make our own.”
Now that the skin cancer identification means has proven largely successful, the team is looking forward to adjusting it in a way that makes it compatible with smartphones.
The scientists hope that they will be able to do so in the near future, as skin cancer presents a significant health risk, with 5.4 million new cases in the U.S. being reported every year.