Artificial intelligence in medicine: translates and determines the age
More than 28 million foreign tourists visited Japan in 2017, which is three times higher than a decade ago. As a result, more and more foreign patients are in the country’s hospitals. As their numbers increase, Japanese companies are involved in the development of multilingual products and services to help hospitals cope with the inevitable language barrier.
Medical startup ShareMedical plans to introduce the service of a telephone interpreter using artificial intelligence by the end of August 2019. The service allows hospital staff and patients to speak on the phone in their native language, by translating the conversation in real time. The robot identifies the languages spoken by people, translates spoken words into text, then converts the text into voice data that is transmitted to the participants.
AI can handle up to 17 languages and uses standard telephone communication, eliminating the need for special hardware or software. The price of the conversation will be about 18-27 cents, which is much cheaper than a live translator.
At the same time, NEC is already working with Japanese doctors helping them fill out forms when registering a foreign patient. Fujitsu company is also ready to provide hospitals with its translating terminals in three languages – Japanese, Chinese, English by March 2019.
According to a survey conducted by the Ministry of Health of Japan in 2016, more than 60% of medical institutions experienced problems with the treatment of foreign patients. Unlike large urban hospitals, it is difficult and expensive for small clinics to attract multilingual staff. As the problem becomes more acute, competition for the provision of language services for health facilities is expected to increase.
Artificial intelligence in medicine determines age by eyes
The researchers introduced a new AI system that can predict age just by looking at a photo of the face. And, that may seem unexpected to many, experts say that the corners of the eye is easiest to determine how old the patient is.
The new technique, tested on more than 8,000 photos, can estimate age with great accuracy, up to three years, even if only a small part of the face is visible.
In their discovery, researchers from HautAI OU skin care products used 8414 high resolution anonymized images to verify the accuracy of their system. They found that the skin near the corners of the eye and the general area of the eyelids were most obvious in predicting age.
This type of information can be used to assess various factors that affect aging, including lifestyle, medical, and cosmetic interventions.