Seth Augenstein

home

authors

seth-augenstein

Francois Bolduc (second from right) and his University of Alberta team are helming a project to develop an AI chatbot for families with children who have neurodevelopmental disabilities. (Courtesy: University of Alberta)

A mother doesn’t know what to do. Her autistic son is having severe difficulties, leading to dramatic problems at home. But they live in a rural part of western Canada, and a specialist appointment is still weeks away.

Soon this mother will be able to ask for immediate help — courtesy of an artificial intelligence (AI)-powered chatbot offering proven advice from the literature.

The AI machine will be able to guide the mother and son in their most crucial time to the best way to deal with the dynamic domestic situation, according to the developers, led by François Bolduc, M.D., a pediatric neurologist at the University of Alberta.

While a growing number of healthcare chatbots are proposed and being built, this one is dealing with a level of human sensitivity surrounding human development which may set it apart.

“That’s what’s kind of unique about this project, many projects are born from AI, but do not necessarily have tight connection with the health of the education sector, which is kind of a very strong thing in my mind, because it gives us this qualified way of processing the questions of that mom,” said Bolduc, in a recent interview with 

The development of the chatbot is underway through a three-year, $1.5-million grant from the Canadian Institutes of Health Research and Natural Sciences and Engineering Research Council of Canada, as well as a development grant from Kids Brain Health Network. Joining Bolduc and his team will be those from the University of Calgary, McGill University, and York University.

Key to the concept is providing resouces as early as possible — resources offered at the clinic, but which can now be accessible from the comfort of the home, said Bolduc.

The guidance would point to vetted Internet sources, with medical and behavioral advice that is tried and true from previous cases.

Although some skepticism of the burgeoning healthcare chatbot market 

, the timeliness and sourced data could make a huge difference to get families in a difficult domestic situation the help they need, at the moment its most needed.  

“A chatbot is perfectly suited for that. It’s there 24/7, you can access it when you have time, you can quit it and come back to it,” he said. “This is really trying to improve quality of life and improving functionality. It’s not going to change the individual… but I think it will make the world a little bit easier for individuals with neurodevelopmental disabilities to live in.”

IBM Develops Virtual Assistant to Aid Stroke Patients

Articles

Chatbot is just the latest of the machines tasked to help patients.

University of Alberta AI Chatbot to Help Families of Kids with Neuro Issues

The pathologists take a sample of tumor tissue, and sequence its entire genetic code. The advanced computer program helps the oncologists process the results, homing in on certain mutations among a group of 400 targets. The genetic hallmarks of the disease are compiled in a text file that is analyzed and transmitted securely by artificial intelligence.

Ten to 15 minutes later, the computer has pulled a review out of millions of pages from the scientific literature — producing a selection of abstracts and pending trials.

From there, the battle against the cancer at the molecular level can proceed, after a consultation of a Tumor Board at the Les Hôpitaux universitaires de Genève (HUG) in Switzerland.

The use of Watson for Genomics, the search engine developed for IBM, has now been used for a dozen cancer patients at the Swiss health system since Aug. 1, becoming the first European site to use the technology.

The oncologists and specialists envision using it for 250 patients per year. While it has been in use in the U.S. and Asia, this brings a new level of automation to the traditionally manual searches conducted by treating oncologists in Europe, according to IBM.

“Already in use in the U.S. and Asia, HUG is the first university hospital in Europe to adopt Watson for Genomics and we are excited to work together to accelerate personalized, evidence-based care for its patients,” said Mark O'Herlihy, IBM Watson Health, managing director, EMEA.

Two institutions that use the Watson technology in the U.S. are Mayo Clinic and the Massachusetts Institute of Technology, according to a statement on the new deal.

Watson for Genomics is being touted for its security aspects — namely that the information transmitted includes no identifying information on the patient, and that the data are deleted at the end of the analysis, according to the experts.

IBM Watson Health says Watson for Oncology (separate from the Watson for Genomics) includes access to 29 million abstracts looking for 13 cancer types, 

according to some of its latest data sheets

. Watson for Oncology’s capabilities include the ability to read 200 million pages in a mere three seconds, according to the company.

The IBM Watson Health team also reportedly debuted 22 new scientific studies at the American Society for Clinical Oncology annual meeting in Chicago, back in May.  

™ in a statement that the process of getting AI into precision cancer medicine is exacting — and is not overnight.

"Adopting AI into healthcare takes time and scientific rigor — we are working closely with hundreds of hospitals and health systems around the globe to bring our oncology and genomics offerings to physicians and patients.”

Get the best insights inside digital health. 

IBM, Aetna, Anthem & Others Announce Collab to Create a Blockchain Network

What IBM Watson Health Hopes to Accomplish Through $50M in Deals with Brigham & Women’s, Vanderbilt

IBM Watson Health Releases Diabetes App, Touts Power of Data

First European use of technology joins counterparts in U.S., Asia.

First European Site Adopts IBM Watson Genomics for Cancer

, our hosts Tom Castles and Jack Murtha talk FHIR (pronounced “fire”) with the CEO of HL7, Chuck Jaffe, Ph.D., M.D.

The Fast Healthcare Interoperability Resources (FHIR) is a draft standard created by the non-profit Health Level Seven International (HL7). It’s intended to further interoperability in healthcare. Founded in 2014, it has incrementally brought together millions of records, and continues to do so, even now.

FHIR has been expanding, especially recently, according to Jaffe’s recent conversation with

“This is growing over at least the last year and a half so that organizations and health systems around the world have leveraged the FHIR platform until it drives interoperability,” said Jaffe. “I suspect that the mass of institutions and entities that have enabled FHIR will have to grow once that mass has been achieved. I think the ability to reduce the barriers to interoperability will accelerate.”

The FHIR Accelerator program was launched, and it recently began incorporating the 

Gravity Project and its social determinants of health (SDOH) into the mix of digital variables

But the real hurdles that remain to true interoperability are not technological, Jaffe said.

“Policy always trumps technology, and we'll find that regulation, inertia and other concerns will continue to be a barrier toward thorough interoperable data exchange,” he said. “The key that is so important in our community is being able to understand what the words and concepts mean. And that development is accelerating in parallel to the data exchange environment in which FHIR is now expanding its use and capability.”

Take the temperature of FHIR. Check out the fifth episode of the fourth season of your favorite digital healthcare podcast, above, or on iTunes.

Navigate the digital transformation with confidence. 

Data Book Podcast: Rediscovering Healthcare's 'Why' With Rasu Shrestha

Why Health Systems Should Build Their Own AI Models

This week’s guest: Chuck Jaffe, Ph.D., M.D., the CEO of HL7.

Data Book Podcast: Talking FHIR with HL7

President Donald Trump signs AI executive order in Oval Office in February 2019. (Courtesy of the White House)

Nearly $1 billion in artificial intelligence (AI) research funding was announced by the White House this week.

The request for non-defense AI funding coincided with an AI summit hosted by the Trump Administration, featuring two healthcare-specific “case studies” of how the future use of American AI is implemented.

“Today, we focus our attention on how all of these pillars of our American AI Initiative can also drive transformation within the federal government itself,” said Michael Kratsios, the chief technology officer of the United States. “Initiatives in AI R&D, data, and workforce can all enhance the way government functions and delivers services.”

 includes $202.5 million for the National Institutes of Health (NIH), the second largest total. (The largest is the National Science Foundation, at $487.6 million). Mentioned in the NIH goals are research project grants, and specifically mHealth programs for low- or middle- income countries.

Also included are Agency for Healthcare Research and Quality (AHRQ) programs including learning health systems grants, health IT safety grants, consumer IT grants for patient-facing technologies, clinical-decision support tools, and a patient-reported outcome initiative (based on the use of Fast Healthcare Interoperability Resource standards).

The summit featured two healthcare case studies in the use of AI from the NIH and from the Department of Health and Human Services (HHS).

Patricia Flatley Brennan, Ph.D., R.N., the director of the National Library of Medicine, spoke about the use of natural language processing at NIH. The AI technology saves time in the grant application referral process, she said. Additionally, AI and machine learning can automate indexing for PubMed.

Charles Keckler, J.D., the associate deputy secretary of HHS, presented how the use of AI can identify and cut “outdated, burdensome, and duplicative regulations, with the potential of greatly reducing human workload in a highly labor-intensive process,” according to 

“The Department of Health and Human Services is using artificial intelligence to analyze outdated, burdensome, and duplicative regulations—allowing us to more effectively cut the red tape holding our nation back,” added Kratsios, in his remarks.

The summit also included a panel featuring Ivanka Trump, advisor to the president, and Margaret Weichert, the deputy director for Management at the Office of Management and Budget and Acting Director of the Office of Personnel Management.

Breakout sessions followed, including federal agencies ranging from the Department of Homeland Security, the Department of Labor, the Department of Justice, and the National Science Foundation, among others.

The nearly $1 billion is not enough to keep pace with the investments of China and other countries, 

Trump signed an executive order in February which created the “American AI Initiative” with a stated goal of “protecting our AI advantage.”

AI Identifies Risk of Fatal Heart Attack 5 Years Before the Event

The Hype and the Upside of AI in Healthcare: A Conversation

Earmarks include healthcare programs at NIH and AHRQ.

Seth Augenstein

Articles

University of Alberta AI Chatbot to Help Families of Kids with Neuro Issues

First European Site Adopts IBM Watson Genomics for Cancer

Data Book Podcast: Talking FHIR with HL7

White House Asks for $1B for Non-Defense AI, Including Health