Hospital IoT highlighted by AI use cases and wearable devices
Healthcare is an ideal landscape for the proliferation of AI. It’s a complex environment that provides copious amounts of data and is technology-rich, making it IoT-friendly. AI driven by healthcare’s rich data, coupled with streams of user-specific data gathered by IoT, can enable better predictive analytics, enhanced outcomes, faster detection of illness and heightened patient self-awareness.
Despite being ripe for AI adoption, healthcare industries face a steep learning curve and a shortage of data scientists. Healthcare is teeming with data but is rife with regulation and compliance issues. With the addition of IoT introducing entirely new complications in infrastructure and integration, deploying AI in healthcare can be daunting — but show promise if done right.
Patients adopting digital healthcare
AI requires a digital healthcare domain to thrive, and the adoption of digital healthcare is about more than new software and infrastructure and methodology — it’s about cultural change.
The biggest of these changes is getting the new tech to the patient. Personal IoT, or wearables, represents the biggest step forward toward making this culture change. Fortunately, most of those devices are unobtrusive and comfortable — like the smart watch.
According to Statista, the number of wearable device users in the U.S. has grown from 32 million to 57 million users between 2015 and 2019. Even so, clear standards for flexibility and interoperability among such devices have yet to emerge. Moreover, IoT itself throws up significant hurdles. Its security paradigm is more complex than cloud technology and traditional on-premises infrastructure, and universal standards haven’t yet come into focus.
Motion devices and AI
Integrating AI to collect more relevant data has improved understanding of patient behavior and emotion, over and above vital signs and treatment plans. When a patient’s actual behavior patterns and feelings enter the analytics mix, outcomes improve.
In the past, getting hold of this extra information was a matter of survey and self-report, spotty collection methods at best. IoT offers a huge advantage here, presenting technology (wearable and otherwise) that can monitor patient activity in real time, collect the data and even serve as an early warning system.
For instance, smart watches now have activity detection functionality. They can tell when a person is running or lifting or exercising and report that activity to an app. The watch can generate activity histories that then can be used to fine-tune a health maintenance regimen or catch a developing condition early on.
Such functionality, however, is based on the machine learning what the motions the device detects indicate. That machine learning requires extensive training means that users of smart watches need to be turning over their behavioral data to the vendors of the watch so that they can be improved over time, which raises privacy and security concerns.
Beyond patient-based IoT, IoT in hospitals themselves is quickly evolving into its own brand IoMT, the internet of medical things.
At a stroke, this variation solves most of wearable IoT’s problems. It’s already centralized, with a single infrastructure pulling in data from all the devices in the environment; though distributed, the IoT network is under a single security/administrative model and persuading the patient to make use of the technology isn’t an issue.
Hospital IoT covers a huge amount of territory, from Bluetooth blood pressure cuffs, fitness devices and thermometers that transport readings directly into the patient’s digital records, to automatic bedside devices that monitor the patient. Then there are the efficiencies and improved outcomes gained from streaming operational data from all patients and processes into machine learning systems. These systems then produce institutional performance optimization through prescriptive analytics.
When hospitals undergo such a makeover, they adopt infrastructure that improve operations beyond their walls — blockchain, for instance, which secures patient data to the point that seamless exchange between providers is made more practical, promoting interoperability, while bolstering HIPAA compliance.
Building better doctors
The deployment of AI and IoT in healthcare diagnostics is a huge focus for hospitals. AI can read an X-ray far more accurately than any human doctor because of its ability to learn quickly, efficiently and process data. While a human doctor might read tens of thousands of X-rays over a career, an AI can study tens of millions. The same applies to medical imaging, and the accuracy of predictions based on it.
The rapid progress of AI’s predictive potential bodes well for future pandemics, where AI-based monitoring of disease vectors and outcomes across geographical areas will underscore the crucial importance of strong data collection. Similarly, AI-assisted surgical robots, while still somewhat rare in the U.S., are swiftly becoming the standard in China.
AI, big data, healthcare and IoT are a complicated mix — but one with the potential to completely transform the game.
Discover Past Posts