This article was originally published on MedTech Intelligence.
The use of artificial intelligence in healthcare is accelerating rapidly, as devices and products that mimic human learning and behavior are on track to get regulatory approval in the United States, Europe and increasingly, Asia.
As recently as one year ago, artificial intelligence was still largely in the research phase, but now hundreds of medical AI applications are nearing approval by regulators. In Europe and the United States, more than 100 AI products have already been cleared for clinical use. A flood of approvals in Asia are not far behind, with some already issued in South Korea, China and Japan.
Drawn to the new technologies because of their promise for speeding diagnosis and treatment of chronic diseases and growing the efficiency of medical delivery systems, the Asian giants are moving fast to get a handle on the new technologies. China’s regulatory authority last year issued its first technical guidelines for AI-based software products, and the government is investing vast sums in development of the new technologies.
AI technologies rely on deep learning and sophisticated computer data analyses to detect disease early, achieve greater accuracy in diagnoses and offer more personalized medical treatments. They are developing rapidly across disciplines, but have been particularly in vogue in radiology, where the use of computerized images and software is already well established. AI technologies can harness large troves of data and image recognition software to provide solutions for image analysis and diagnoses. AI algorithms are being used to interpret chest radiographs, detect cancer in mammograms, identify cancerous skin lesions and colon polyps, analyze computer tomography scans, spy brain tumors on magnetic resonance images and even predict the development of Alzheimer’s disease from positron emission tomography. Other applications of AI rapidly coming into play include drug development, health management and analysis of electronic medical records and genetic markers.
In China, AI is Deployed against COVID-19, and More Widely for Disease Prediction
Chinese health officials and private companies are seeking to leverage artificial intelligence to solve the nation’s most pressing healthcare conundrums—a dearth of medical resources, high costs, arduous training periods for physicians, and an elevated rate of faulty diagnoses. Since the outbreak of the COVID-19 virus, AI is being marshalled to address all those issues right away.
At the end of January, the Shanghai Public Health Center (SPHCC) and Yitu Healthcare, a Shanghai-based startup, launched what they call the Intelligent Evaluation System of Chest CT for COVID-19. It uses AI technology to improve on traditional methods to evaluate chest lesion images, which usually require several hours. While those traditional methods are plagued by low efficiency and difficult clinical promotion, Yitu says the new AI system marshals image algorithms to intelligently diagnose and quantitatively evaluate CT images of COVID-19. It rapidly grades the severity of local lesions, diffuse lesions and whole lung involvement, allowing complete quantitative analysis in seconds.
The SPHCC is also combating the spread of the coronavirus using another AI technology, a continuous temperature sensor developed by VivaLNK, a Santa Clara, California-based AI startup. And in mid-February, the Chinese government released a public app, also AI-powered, to gauge potential coronavirus exposures.
Chinese multinational tech company Baidu, Inc. has made its online doctor consultation platform free for users who want to consult with a doctor about COVID-19. Thus far, the platform has handled more than 4.2 million inquiries from users about COVID-19, according to Baidu. And in hospitals in several Chinese provinces, AI-powered robots are being used to boost staffing and reduce the risk of cross infection among nurses, doctors and technicians
Beyond AIs application to the COVID-19 outbreak, Chinese firms are moving forward in using AI to improve medical imaging and diagnosis over the long term.
Since 2017, when Shenzhen-based-based Tencent Holdings, Ltd. launched its AIMIS diagnostic medical imaging service, it has reached agreements with more than 100 Chinese hospitals to employ the technology. Airdoc, a Chinese health technology company, has developed an AI-driven system that diagnoses chronic diseases by analyzing retinal images. Also, Accutar Biotechnology, a Chinese AI pharmaceutical company, has raised millions of dollars in funding for its venture to use 3D projections and chemical structures to develop pharmaceuticals.
Other Chinese companies are working to develop smarter wearable healthcare prediction and diagnosis devices, too. San Diego-based 12 Sigma Technologies, a company with a deep footprint in Beijing, Suzhou and Shanghai, makes data analysis software that allows physicians to diagnose patients quickly by inputting an X-ray or CAT scan image to scout for multiple disease indications at once. And iCarbonX, based in Shenzhen, has invested more than $350 million in developing machine learning algorithms, that company officials say can find patterns in healthcare data to prevent and diagnose disease.
Japan Healthcare Companies Deploy Machine Learning, Other AI Technologies
Healthcare firms in Japan are increasingly developing AI technologies as well. Japan’s MHLW/PMDA has granting expedited review authority to some new AI-based products since the start of 2018.
Among them are endoscopy AI devices to improve detection of colorectal cancers. One Tokyo startup, AI Medical Service, is developing AI technology to detect not just colorectal, but esophageal and gastric cancers, applying neural network algorithms trained on massive sets of data captured using standard endoscopes.
Another Tokyo-based company, Fronteo Healthcare, is developing software it says uses electronic medical records to help hospitals identify older patients with a proclivity to fall and injure themselves.
South Korea, Singapore Firms Arming against Public Health Threats with AI Technologies
The COVID-19 outbreak is also spurring the development of AI technologies outside of China. In Singapore, the government announced in February it is partnering with KroniKare, an AI startup, to pilot smart phone-based temperature screening technology to identify people showing symptoms of fever. The system uses a cellular phone fitted with thermal and 3-D laser cameras to gather and analyze multispectral images and temperature data up to nine feet away. The company says the device can detect people who may have fever as they walk by a subject carrying the device.
In South Korea, AI firms are taking on the growing prevalence of Type 2 diabetes (more than 5 million Koreans have been diagnosed with the disease) by using the power of metadata to manage the chronic condition. Korean company i-SENS is among a number of firms that have developed glucose checking devices that employ AI to speedily gather data on a patient’s condition and transfer it to their physicians.
The Path Ahead for AI: Faster Diagnoses, Greater Efficiency, Rapid Adoption, Better Analysis
In Asia as elsewhere, the promise of artificial intelligence is rapidly coming to fruition. Going hand in hand with the digital revolution, it is harnessing big data storage, advanced analytics and the power of networks to speed diagnoses, trace pharmaceutical use, and make doctors and other healthcare professionals more efficient. If successful, the breakthroughs will forge a new healthcare ecosystem where AI technology is the first line of preventative care and treatment, and a key partner in keeping people healthy.