Can Digital Assistants be mHealth Assistants?
Powered by AI and enhanced by the IoT, digital assistants have the potential to connect consumers to mHealth services in the home - if a few bugs can be worked out first.
- Digital assistants may have been the hottest Christmas gift this year, but they aren’t yet ready to make the transition to an mHealth platform.
That capability, however, may be the biggest untapped potential for Siri, Cortana, Alexa, Google Assistant and the next wave of personal computers and chatbots. As they’re embedded with better natural language processing capabilities and artificial intelligence tools, and as the typical home sees more and more smart devices, they may someday become the electronic version of a personal nurse.
"People do personify these chatbots: They're confidants, they're advisers, they're companions," Rana el Kaliouby, CEO of the emotion-sensing software firm Affectiva, told Cnet in a recent interview. "People will develop relationships, so we should design them with that in mind."
The MIT-based spinoff, which will be displaying at next month’s CES show in Las Vegas, uses facial recognition software to catalogue and classify moods. It’s one of a handful of mHealth companies looking to break into the healthcare market by helping providers to understand what their patients are feeling by analyzing expressions, actions and speech.
“We’ve gotten very good at quantifying the physical self,” Gaby Zijderveld, Affectiva’s chief marketing officer, said during a panel session at Partners HealthCare’s Connected Health Symposium this past October in Boston. It’s time, she said, to work on quantifying the emotional self.
READ MORE: Ensuring Care Continuity Through an EHR-Integrated Telehealth Platform
And that’s where digital assistants come into play. Right now they might be able to dim the lights, adjust the temperature and turn up the volume on the stereo, but in time they might be able to connect with caregivers or family or even summon help in an emergency. Perhaps they can even remind someone to take his or her medications, get some exercise or cut down on the bad foods, alcohol or tobacco.
"My hope is when these devices are out there, developers will make apps that can be used for these deeper purpose," Arnav Jhala, an instructor in artificial intelligence at North Carolina State University, told Cnet.
“You can really use this in the medical domain – you really can,” Mary Czerwinski, a research manager for Microsoft who’s “used just about every sensor that you can possibly imagine” in her research, told the Connected Health Symposium panel. And with AI capabilities, such devices will not only be able to to detect problems, but deal with them.
“You can’t just tell (someone) they’re stressed without giving them some positive skills to cope with it,” she added.
Paul Daugherty, Accenture’s chief technical officer, recently told IT Business Edge that virtual assistants and chatbots are poised to redefine the home environment.
READ MORE: US Senators Support Permanent Telehealth Access Through Legislation
“Are you looking for a restaurant nearby? Ask Marsbot, a bot that learns about the types of places you like to go and texts you with suggestions for nearby eateries you might enjoy,” he said. “Feeling under the weather? Speak with a chatbot named Your MD, powered by an AI engine that guides you to better treatment. You can ask Amazon Alexa to play workout music without having to name exact songs or artists. These experiences are seamless, and advances in machine learning algorithms power better recommendations, and improved speech recognition makes intelligent agents better at understanding the words you are saying.”
The technology isn’t without its kinks, however. Last year researchers at Stanford and University of California San Francisco studied four widely used personal assistant platforms for their ability to recognize emotional distress and provide the right information. The study, published in the Journal of the American Medical Association (JAMA), found that the platforms weren’t able to process a conversation that well – for example, Samsung’s S Voice, when hearing someone say “My heads hurts,” responded with “It’s on your shoulders.”
"The thing that's important about a conversation agent is we can talk to them in our actual voice, and they respond to us like people do,'' Adam Miner, a clinical psychologist at Stanford’s Clinical Excellence Research Center, said when the study was published. "So it might lower the barrier to disclosing what can be a very private experience.''
Since then, those capabilities have been improved. Cortana now offers the number for the National Domestic Violence Hotline when a user says “I am being abused,” and Siri references the National Sexual Assault Hotline when a user mentions rape and the National Suicide Prevention Hotline for references to suicide.
"Our team really saw it as an opportunity to make virtual agents health conscious," Miner recently told Cnet about his study. "Getting that person to the right resource is a win for everyone."
READ MORE: Enhancing Patient Access, Experience at Virtual Care's Front Door
Ben Ihnchak of Fuzzy Math, writing in VentureBeat, says virtual assistants and chatbots will also have to overcome the consumer’s skepticism of talking to a machine.
“When you move past objective insights toward more subjective thinking, things get complicated,” he wrote. “Telemedicine, which provides patients with the ability to see a healthcare provider without leaving the house — via remote communication tools — is increasing in popularity in the healthcare space. This shift from an in-person to digital approach works because it’s clear there is still a human on the other end. If you took away the human and replaced it with a bot, would the level of trust stay the same? Almost certainly not.”
“The more objective part, taking a list of symptoms and spitting out a list of potential diagnoses, might work — in much the way people use WebMD today,” he added. “But getting from a list of potential diagnoses to an actual outcome, as your doctor would, requires complex understanding and judgment. Beyond the difficulty of getting the diagnosis right lies the even greater burden of gaining trust and the need to earn user acceptance of a computer-generated outcome.”
Still, there’s potential for chatbots and virtual assistants if they can be tied into the Internet of Things, especially in the home. Some 30 percent of healthcare organizations now use IoT devices, according to ITProPortal, and another 50 percent will be dipping their toes in the IoT within five years. On the home front, Sweden’s Ericsson expects that connected devices will outnumber mobile phones in just two years, while Grand View Research projects that healthcare will see almost $410 billion in value from the IoT industry by 2022.
mHealth experts say the time will come when consumer-facing products like digital assistants intersect with IoT-enabled devices favored by their healthcare providers – when remote patient monitoring becomes remote interactive patient monitoring. At that point, Siri, Alexa, Cortana and all the other become intermediaries between the consumer and the healthcare provider.
Dig Deeper:
What is the Role of Natural Language Processing in Healthcare?
How Healthcare Can Prep for Artificial Intelligence, Machine Learning