Science

Artificial Intelligence Used To Identify Skin Cancer Effectively

By Rodney Rafols , Jan 25, 2017 11:59 PM EST
Close
Security robot falls into a fountain in Washington DC
MIAMI, FL - JUNE 15: Doctor Antonella Tosti, Dermatologist University of Miami School of Medicine, displays an image on her iphone of a mark on the skin of Michael Casa Nova, 12, as she uses it as a dermatoscope while examinging him for symptoms of skin cancer due to sun exposure on June 15, 2011 in Miami, Florida. The federal Food and Drug Administration announced that sunscreen manufacturers are to change the labels on their products to prohibit the use of certain marketing terms. The new rules are meant to help clear up confusion about the meaning of 'sun protection factor,' or SPF, and other terms like 'waterproof.' (Photo : Joe Raedle/Getty Images)

Computers today are being used more and more in many tasks. Modern technology is moving forward and is taking over what was once was done by humans. Artificial intelligence can now be used to identify skin cancer effectively.

Researchers have created an algorithm that could perform just as well as the best dermatologists in looking at skin cancer. Computer scientists from Stanford University have created an algorithm by gathering a database of around 130,000 skin cancer images. The images were then input into the algorithm so that the computer could diagnose skin cancer.

Sebastian Thrun, adjunct professor at the Stanford Artificial Intelligence Laboratory, said that the algorithm is very much feasible. He also said that the algorithm can not only do well, but do as well as the best dermatologist. The algorithm has been made, as he stated, not just as a class project but as something for humanity.

The computer algorithm has been tested against 21 board-certified dermatologists. In the test it has been shown that the algorithm has performed just as well as the dermatologists did. The algorithm could be used as an additional tool for dermatologists in diagnosing skin cancer.

In diagnosing skin cancer, dermatologists first make an observation of the area affected. This might mean looking for suspicious marks or differences on the skin, according to the Stanford News Service. Dermatologists might also use a small handheld microscope in order to see the skin much closer.

The use of the algorithm in skin cancer diagnosis would combine visual observation along with deep learning. Deep learning is a computer process that has been patterned after the neural brain network. In deep learning, the computer is trained in the task, as Science Daily reports. As the algorithm looks at more skin cancer types then, it can adjust itself in order to know more about it.

The algorithm that has been made is based on the same one made by Google to identify 1.28 million images. In this case, the algorithm has been revised in order to identify skin cancer. The research team hopes to make the algorithm be available on smartphones as well. This would make skin cancer diagnosis be more accessible to more people.

Skin cancer diagnosis is made easier with a new algorithm designed to identify it. Artificial intelligence can be used to identify skin cancer more effectively. A study has also shown that diet soda could also give diabetes.

© 2017 ITECHPOST, All rights reserved. Do not reproduce without permission.
Real Time Analytics