Happy New Year everyone! A warm welcome also to my Fortune Jonathan Vanian, partner of “Eye on AI”, who has just returned from paternity leave. Jonathan helped compile the news, research and Brain Food sections for this week’s newsletter.
I’ve spent some time over the past few weeks reporting an article, soon to be published, on how businesses hope to use AI to help COVID-19 vaccination campaign. The global vaccine rollout, which is now underway, although progressing much slower than most would like, will be among the most important stories of this year and it is one of the can play a role in at least four distinct areas:
• ˘Triage and impact modeling. Determine which population groups should be vaccinated in order to end the pandemic quickly.
•Demand forecast. Determine where and when to ship the doses to get so many people vaccinated in the shortest possible time.
•Supply chain management. Monitoring of the vaccine production and distribution network for bottlenecks.
•Post-vaccination surveillance. Watch for signs of unwanted side effects of the vaccine that may not have been detected in clinical trials.
Today I want to talk about just one aspect: demand forecasting. In my report, I spoke to Benjamin Fels, who is the co-founder and CEO of a small business in Boston called Macro-eyes. Fels is an intriguing character. With pale skin, piercing blue eyes, a receding hairline and a bushy black beard, he reminded me of a 19e missionary of the century. And he’s sort of on a mission. Its goal: to improve health care, especially for those in low- and middle-income countries.
Fels first discovered machine learning while working at a hedge fund in Chicago and London. He was intrigued by the idea that an AI system could be fueled by many disparate datasets and, by detecting a few weak signals in each of them, then putting together robust predictions that traders could use to win. money in the financial markets.
But Fels just didn’t want to do well. He wanted to do good too. And he wondered what the powerful machine learning systems he was using could do if applied to different types of data. “There must be, I thought naively, other areas where this is important,” he said.
He teamed up with Suvrit Sra, a computer scientist at MIT, and together the two decided to take on healthcare. “Our basic assumption was that health care, the practice of medicine, the whole question of how you provide health care, is a protracted act of pattern recognition,” says Fels.
Their first client after founding Macro-Eyes was Stanford University Healthcare. Like many health care practices, one of Stanford’s biggest problems was not showing up – patients who didn’t show up for appointments. Macro-Eyes has built a system that can reliably predict which patients are likely to miss their appointments. Critically, it could also suggest alternative clinic times and locations that would increase the chances of a patient presenting.
The company has since applied similar technology in locations from Arkansas to Nigeria, with support from the U.S. Agency for International Development and the Bill & Melinda Gates Foundation, among others. “We’ve become very good at predicting health-seeking behaviors,” says Fels.
So what does this have to do with the COVID-19 vaccine? Well, in Tanzania Macro-Eyes worked on a project to improve childhood immunization rates. Its software analyzed the number of doses of vaccine to send to each vaccination center. Critically, he also identified families who might be reluctant to vaccinate their children but could be persuaded with the right message and a conveniently located vaccination center. Using this software, the Tanzanian government was able to improve the efficiency of its immunization program by 96% and reduce wasted vaccine doses to 2.42%.
The lessons of Macro-Eyes in Tanzania are directly applicable to the COVID-19 vaccination effort, Fels says. Ensuring the effectiveness of this vaccination program is critical as demand is likely to exceed global supply for some time and, especially in low-income countries, every dose will be valuable. Concerns about cold storage will be an issue in many places, even with vaccines such as AstraZeneca that can be stored at normal refrigerator temperatures. “We can’t afford to throw away 30 doses because we sent them to the wrong place,” Fels says. “But if we allocate the vaccine strictly on the basis of population, we will severely under-allocate to some sites and severely over-allocate to others.
To understand demand, he says, Macro-Eye’s artificial intelligence software examines datasets that include geospatial data, especially satellite imagery, information about the number of mobile phone users in an area. data, government and health system records if available, and sometimes even social media posts.
Businesses are used to having to be creative with the data they use. In Sierra Leone, which lacked health data, the company had to figure out how to correlate information on local school infrastructure to make predictions about the quality of local health centers. It turned out that the teacher-student ratios alone were enough to predict with 70% accuracy whether the local clinic had access to clean drinking water, Fels says. “Each of these data sources alone is usually of marginal value,” he says. “But if you combine them, you start to get an accurate picture.”
This is great advice for using AI in business, as well as an example of using AI in business to do good. Hopefully the New Year will bring a lot more. And with that, here’s the rest of the AI news from this week.