Artificial Intelligence in Medicine

by Robin Farmanfarmaian

In 2016, I wrote about some of the most exciting and promising AI developments of the year. A DeepMind artificial intelligence program successfully won four out of five Go matches against a renowned master, in the process showing its ability to learn. With the promise of eliminating more accidents and making daily commutes more pleasant, self-driving cars continue to make forward progress — with an increasing number of corporations in the development of driverless vehicles.

The most impressive advancements in artificial intelligence may have been in the field of medicine. IBM’s supercomputer, Watson, turned its AI talents toward genomics with a focus on cancer and amyotrophic lateral sclerosis (ALS). Backed by Bill Gates and Jeff Bezos’s investment firm, biotech company Illumina began work on the Grail, a single blood or urine test designed to identify any form of cancer in an earlier, treatable phase.

Artificial intelligence has continued to make impressive breakthroughs in medicine throughout 2017. From companies taking advantage of IBM Watson’s open API to big names like Apple and Amazon — as well as rising startups — medical AI continues to be a fast-moving field inspiring developments with wide-ranging benefits. The following highlights just a few of the most promising projects I’ve discovered in medical artificial intelligence.

Building on IBM’s Watson
Pharmaceutical giant GlaxoSmithKline is using Watson to provide a more personalized experience to customers. The system is able to give GSK customers individualized attention by providing customized responses to text and voice questions. This real-time assistance provides far better suggestions than what these customers could find on their own with a Google search.

At the American Cancer Society, Watson is being used in a similar way — providing detailed advice and information to cancer patients. The power of IBM’s AI allows patients to ask questions verbally and receive audio responses. This virtual adviser leverages data from Cancer.org, the National Cancer Information Center and IBM’s Watson Health Cloud to offer the most complete responses possible.

For the organization known as Health&, IBM’s Watson powers an online content library of evidence-based information. It also manages a “popular questions” page, keeping the most relevant and timely answers within easy access. Watson Health and Pfizer Inc. have partnered to expand on Pfizer’s immuno-oncology research, which uses the body’s own immune system to fight cancers. As an early adopter of Watson for Drug Discovery, Pfizer intends to use the AI tools to identify new drugs, combination therapies and patient selection strategies.

Key Brands Investing in Medical AI
Apple has seen the benefits of getting in on the field of medical artificial intelligence, recently acquiring several startups with a healthcare influence. Its ResearchKit has already been used to collect data on health conditions such as asthma attacks, seizures and heart disease. Not only does ResearchKit provide real-time information to Apple Watch wearers, allowing them to adjust to immediate health needs, but this information is also providing a basis for further research by scientists.

Combining medical AI with popular smart home assistant technology, Amazon’s Alexa recently added support for HealthTap’s Dr. AI. This program lets home users ask questions and find out more information on health issues they are experiencing with voice activation, a huge advantage for those who struggle to use a more traditional computer interface. Using its database of medical information, Dr. AI identifies the most likely cause of symptoms and can connect the user with a HealthTap-associated physician when needed.

Microsoft is also looking to make a mark in the field of healthcare with artificial intelligence. Working with the University of Pittsburgh Medical Center, the personal computing company seeks to assist physicians with easier access to electronic health records and other digital healthcare systems. In turn, this can improve productivity and allow healthcare professionals to assist more patients, without getting burned out.

Medical Artificial Intelligence in the Startup Space
Enlitic is a technology startup that leverages artificial intelligence’s natural skill in pattern recognition. Deep-learning technology gathers and analyzes data in order to interpret information and solve more unusual medical cases. In recent testing, it was 50 percent better at identifying malignant tumors than radiologists and had a perfect score when it came to not missing cancer (compared to 7 percent false negatives for human radiologists).

In San Francisco, 3Scan is using robotic microscopes and machine vision to create better views of tissue samples while reducing the amount of time it takes to complete meticulous tissue analysis processes. For comparison, the 3Scan machine can complete a tissue analysis in just one day, where the same analysis would take up to a year for a traditional pathologist.

iCarbonX, a China-based startup, is focused on making digital versions of individuals based on biological samples along with environmental and lifestyle factors. Based on this information, its plan is to provide customers with personalized recommendations for wellness practices, food choices and even medications.

Conclusion
With revenues estimated to be $6,662.2 billi on in 2021, medical artificial intelligence is still on a significant growth path as we look toward the future. From facilitating more convenient and accurate self-diagnosis to analyzing data from complex medical systems, the field is in an exciting transition period that is sure to bring even more breakthroughs in the years to come.