Release date: 2017-02-13
Skin cancer is one of the most common malignancies in humans. In theory, they can be easily found on the surface of the skin, but people tend to ignore them because they are too similar to skin cancer. When the condition deteriorates, it is too late.
This also blames ordinary people. Identifying skin cancer is a complex task . Doctors need to analyze its shape and characteristics, and even use biopsy techniques to cut a piece of tissue from the patient's tissue to determine if the foreign body on the skin has actually become cancerous. This is obviously too complicated for ordinary people without medical knowledge.
Last week, a paper published in the journal Nature gave us a handy tool. Behind it is deep learning and artificial intelligence that are not so popular now.
â–² In the eyes of artificial intelligence, skin diseases are like this (Source: Nature)
As we all know, the learning ability of machines is far from being comparable to human beings. In this study, scientists asked a Convolutional Neural Network to analyze nearly 130,000 clinical skin cancer images, two orders of magnitude higher than the largest research available today. With a large amount of learning materials, this neural network quickly became an expert in skin cancer.
It is said that practice is the sole criterion for testing truth. This neural network is not allowed to be accurate, and it is only tested by practice. Together with it, there are 21 senior dermatologists. A "human-machine war" in the medical field is on the verge.
â–² Different skin diseases that seem close to each other, but in the eyes of artificial intelligence, they have completely different meanings (Source: Nature)
In the first challenge, this neural network worked with doctors to distinguish between two different skin diseases, keratinocyte cancer and benign seborrheic keratosis. The former is the most common skin cancer. This comparison found that the neural network performed better than most of the dermatologists involved in the study, in terms of comprehensive sensitivity and specificity .
The unconvinced humans did the second test, this time they compared malignant melanoma with benign sputum. The former is the most lethal skin cancer. But in this test, humans are also defeated.
More exaggerated is that human doctors need more than ten years of training and training, and they need to be passed down from generation to generation. Artificial intelligence has no such restrictions at all. In fact, it will only get better and better.
â–² This study is on the cover of Nature (Source: Nature)
What does this mean for humans? It is estimated that by 2021, there will be approximately 6.3 billion mobile phones in the world, almost one person. If you install the analysis program on these phones, it is equivalent to 6.3 billion professional dermatologists. At that time, everyone can easily take pictures, scan and analyze the foreign objects on their skin, and know their risk at the first time.
An exciting era of artificial intelligence is coming!
Source: WuXi PharmaTech
Professional Super Food Ingredients manufacturer is located in China, including Barley Grass Powder,Spirulina Tablet,Malt Barley Powder, etc.
Our healthway supply Medical Intermediate Herbal Extracts Natural Plant Herbal Extract Super Food Ingredients Fruit Powder Fruit Juice Powder Freeze-dried Powder Vegetable Powder Food Additives Cosmestic Ingredienrs
Icariin NMN powder Cissus Irish sea moss Quercetin Barley grasss, We had passed FDA certificate and ISO 9001 .
Super Food Ingredients,Barley Grass Powder,Spirulina Tablet,Malt Barley Powder
Xi'an Healthway Biotech Co.,Ltd , https://www.xianhealthway.com