By Cory Kidd, Catalia Health
Artificial intelligence (AI) is impacting every industry, with exciting implementations resulting in improved consumer experiences, enhanced technological capabilities, and increased efficiency. Healthcare is no exception.
The market for AI in this sector is expected to hit $6.6 billion by 2021. With new applications emerging at an incredible pace, the technology is disrupting the field across patient care, pharmaceutical research, medication adherence, and countless other areas. One notable application where tremendous growth is taking place is that of device development.
The paramount goal of applying AI towards healthcare technology is to achieve the ability to better adapt, learn, and optimize the outcomes for each patient. This enhanced tailoring to patient need heralds enormous levels of increased efficiency and customization that are difficult, if not often impossible, to achieve without such technology. Given the level of regulation in the sector and the significant responsibility of assisting in patient care, it’s important that innovators, providers, and other parties involved have a high-level understanding of both the benefits and the limitations of AI technology.
How AI Is Overhauling Medical Device Design
Medical devices are swiftly transforming beyond the traditional concept of a single device that is designed, manufactured, and delivered, and then never evolves further. Today, we’re witnessing the advent of devices that incorporate machine learning and AI algorithms that help them evolve intuitively, without human involvement after deployment. These algorithms can actually learn something about a particular patient, and then adapt to that patient’s individual needs in real time. Devices like hearing aids and biosensor monitoring tools are being reimagined by incorporating AI algorithms.
Newer devices to the market, such as wearables that go beyond simple activity trackers, also are gaining traction following widespread adoption of consumer models. Machine learning algorithms in such devices are being customized to target and support specific chronic conditions to help monitor patient systems. The tools also are providing essential data to doctors that can help improve treatment plans and, ultimately, health outcomes.
The first wave of AI-enhanced devices has very recently been approved by the FDA and is now becoming available to patients. A device to autonomously detect diabetic retinopathy, as well as software to assist clinicians in identifying wrist fractures, or to help neurovascular specialists evaluate patients, all have been approved by the FDA this year.
Another trend arising in an array of devices is the inclusion of features for patients to interact with beyond only a button or two. Examples of such interactive tools include advanced patient apps that provide customized advice, applying machine learning to data based on the patient’s condition, symptoms, and other personalized factors. Late in 2017, FDA approved the Reset app from Pear Therapeutics to help treat substance abuse disorders. Another example is found in interactive robots that use AI to tailor a conversation to a particular patient at a specific point in time — these can support treatment and relay important patient data back to providers to improve outcomes.
Why Understanding Psychology Is Essential
As innovators and providers seek to successfully provide this level of interaction for patients, we must first care about and understand the psychological aspects involved in what these patients are dealing with, and then find a way to meet those needs through capabilities available to a device that will interact with that person. This is not a new concept; the fields of human-computer and human-robot interaction have grown equally out of psychology and technology over more than a half century of work. As we explore the new frontiers of AI and machine learning, and how these technologies can affect patients, we need to ensure that psychology is as fundamental a principle as engineering or manufacturing in designing new devices .
When designing devices intended to interact with patients over extended periods of time, there are two types of psychology that we must draw on. The first is the psychology of relationships — in particular, understanding the arc of how a relationship evolves over time. When we start to take this into account, one of the things that we realize is that not every interaction is the same, depending on when in the life cycle of device use this interaction occurs.
The second factor is the psychology of behavior change. One of the biggest challenges we face in creating devices intended for long-term use by patients is supporting the devices’ actual use in daily life. In other words, it's not enough to build something that simply has a medical function. If we can't help support the patient in using it, over time, by creating engaging, user-friendly, easily manageable, intuitive devices, we're not going to succeed in terms of outcomes. It's clear, as evidenced by adherence rates to many currently available treatments, that proper adherence is an enormous challenge for each individual device to overcome
What To Expect Next
What we're seeing today is just the beginning of the use of psychology and AI in device design. Over the next several years, we're going to see a rapid increase in the use of this science and technology to improve the ultimate outcomes we are targeting when creating devices for patients. As the FDA has increased the rapidity with which they consider and approve AI-based devices, we’ll see a plethora of applications across the diagnostics and treatment space.
First will be imaging-related diagnostics, but AI is starting to be used in more complex areas of disease management, and in taking on some of the patient adaptation that has historically been a very manual process. Continuous glucose monitoring and insulin pumps are getting a lot of attention, as well, and behavioral aspects of care will not be far behind.
In June, the AMA recently released guidelines surrounding “augmented” intelligence to help direct design, development, and implementation of the technology in patient care. The organization’s sentiment was that education is key for both clinicians and patients. They also suggested that doctors can work alongside developers throughout device creation to improve real patient implementations, and to provide expertise to avoid potential roadblocks. As we’ve witnessed the vital role that psychology principles play in successful patient-device interaction, I’d suggest it’s vital that developers and providers take these into consideration, as well.
About The Author
Dr. Cory Kidd is the founder and CEO of Catalia Health. Dr. Kidd has been working in healthcare technology for nearly two decades with his work focused on applying innovative technologies towards solving large-scale healthcare challenges. Dr. Kidd received his M.S. and Ph.D. at the MIT Media Lab in human-robot interaction. While there, he conducted studies that showed the psychological and clinical advantages of using a physical robot over screen-based interactions. Follow Dr. Kidd on LinkedIn and get in touch at email@example.com.