AI outperforms dermatologists in diagnosing skin cancer

Share

A CNN is an artificial neural network that is inspired by the biological processes of the human brain's nerve cells.

The artificial intelligence, a so-called convolutional neural network, has managed to be better than dermatologists at spotting skin cancer by simply scanning series of photographs, announced today the research team.

They compared CNN's performance with the performance of 58 global dermatologists.

US says North Korea behind malware attacks
Cho said he remains hopeful for the meeting between Pompeo and Kim Yong Chol, slated to take place later this week. Kim Jong Un is reportedly offering to open an American burger chain in North Korea as a "show of goodwill".

The performance of the CCN was later tested against 58 dermatologists belonging to 17 different countries which revealed that the CNN missed a few melanomas and misdiagnosed a few benign moles as malignant when compared to the dermatologists.

The AI uses machine learning to identify and understand what it "sees" in images, with researchers saying it quickly learned the difference between malignant and benign moles.

Aadi Kalloo, an author of the study and senior data analyst at New York's Memorial Sloan Kettering Cancer Center, contends that "a human definitely always needs to be in the loop" when diagnosing skin cancer, adding that the CNN's diagnostic performance was superior to most, but not all, dermatologists. When provided with clinical information and close-up images of the cases, the dermatologists were able to accurately diagnose 88.9 percent of malignant melanomas and 75.7 percent of benign moles.

Italy's PM-designate Giuseppe Conte fails to form populist government
After the president blocked Mr Savona's appointment, Five Star leader Luigi Di Maio called on parliament to impeach the president. Mattarella's office said it was studying Di Maio's proposal "with attention". "We need to keep cool".

Each year, the scientists said around 232,000 new melanoma cases are recorded with 55,000 deaths due to the disease in the entire world. In the end, the dermatologists were only 86.6% accurate at diagnosing skin cancer, while the computer was able to diagnose issues with a 95% accuracy. "Before doing so, 100 of the most hard lesions were selected to test real dermatologists in comparison to the results of the CNN", he said. Total body photography is used to track potential changes in moles and lesions for early detection of skin cancer. "However, the CNN, which was still working exclusively from the dermoscopic images with no additional clinical information, continued to out-perform the physicians' diagnostic abilities".

"Currently, there is no substitute for a thorough clinical examination", Dr. Victoria Mar, of Monash University in Melbourne, Australia, and H. Peter Soyer of The University of Queensland in Brisbane, Australia, said in an editorial that accompanied a study in the Annals of Oncology, according to a European Society for Medical Oncology statement.

But it is unlikely that a machine will take over from human doctors entirely, rather functioning as an aid.

Trump is back to publicly bashing Jeff Sessions
President Donald Trump says that he wishes he had not chosen Jeff Sessions as attorney general. Government officials don't conduct or oversee investigations that might implicate themselves.

According to the researchers, their deep learning network could one day help dermatologists in screening skin cancer and making the right decision to either biopsy a lesion or not.

Share