Artificial intelligence is already playing a role in healthcare delivery, on the clinical and statistical analysis side. Here, 10 leaders of healthcare and IT companies describe what the future could look like when artificial intelligence is more fully integrated.
Adam C. Powell, PhD. President of Payer+Provider Syndicate (Boston): "We are already taking baby steps towards incorporating artificial intelligence in healthcare. It is already playing a role to some extent, in the form of clinical decision support systems. The use of AI is gradually evolving from providing alerts to assisting in the diagnostic process. AI will at first play a role in making routine diagnostic judgments, and then will gradually be able to be used in low-frequency clinical events. While we often associate AI with voice-based assistants and robots, it is already being used in a simpler form through automated lab test processing and interpretation. Pathology and diagnostic radiology are particularly ripe for AI, as specialists in these fields are fed a digital signal (digital images) and then produce a digital output, in the form of a diagnosis.
In 20 to 30 years, an increasing proportion of routine diagnosis will be able to be handled through AI. Physicians in fields which are not procedural may be particularly vulnerable, in that it may be easier to automate the diagnostic process than the performance of procedures. To prepare for the move towards AI, physicians should focus on acquiring niche diagnostic skills so that they can be useful in situations which are less likely to be automated.
Physicians with niche skills may likewise be useful in helping curate and optimize algorithms used during the diagnostic process. Furthermore, they should invest in enhancing their procedural abilities and patient counselling skills. Even if AI plays a key role in diagnosis, patients will rely upon physicians to help them interpret and adapt to diagnoses."
Kapila Ratnam, PhD. Partner at NewSpring Capital and NewSpring Healthcare (Radnor, Pa.): "In the short term, AI has immediate applicability within the administration and operation of a hospital or healthcare system; more so than in changing how clinical care is delivered. This is predominantly because of how the healthcare ecosystem is regulated; it is one of the most highly regulated industries because it directly impacts individual lives.
Think more in terms of supply chain and inventory management within a hospital; for example, making sure a surgical room is correctly stocked with appropriate inventory based on the series of surgeries to be performed in that room on that day. Having inventory automatically tracked, with an algorithm that is able to assess what the needs are and send instructions to a robot to pick up and deliver appropriate supplies to the surgical room at the beginning of the day. I can also imagine a scenario where an algorithm is able to automatically access cost of supplies from multiple vendors and immediately analyze and assess which vendor delivers the best product at the lowest cost. And follow that up with ordering inventory appropriately, all done without human intervention.
Other applications are in security. For example, assessing whether someone within a health system is inappropriately accessing data and being able to flag that before there is an actual breach of clinical data (there is already a company that has built out an algorithm to do exactly this). Most breaches occur because an insider either deliberately or inadvertently accesses data they should not have."
David Reid. CEO of EaseCentral (San Francisco): "There is no doubt that millennials are a rising buying force, and the healthcare industry is no different. Now is the time for healthcare professionals to build these new relationships and assist millennials in finding their lifelong healthcare professionals of choice.
At the most basic level, the growing telehealth industry has already begun implementing AI into its diagnosis process. Users can upload symptoms, which the makes suggestions to doctors/nurses in real-time. Similarly, doctors can use AI to find the best medication options for a specific patient more efficiently than ever before. AI can play a role in nearly every step of the healthcare system, even acting as a virtual nurse who ensures patients take their mediations (i.e. AiCure), and providing easy-to-manage portal access to discharge instructions and follow-up care. It can even play into insurance, helping users find transparency in pricing and coverage via chatbots and similar consumer-friendly tech tools.
On the employee benefits side, there has been an increase in demand for tech (like telehealth) to be included in benefits care packages. Consumers have grown accustomed to having easy to use, transparent and smart technology in their everyday life, and they expect the healthcare industry to be no different. Healthcare professionals should therefore ensure they have offerings that match this rising shift in demand."
Brian Sanderson. Managing Principal of the Healthcare Services Group at Crowe Horwath LLP (Chicago): "Artificial intelligence provides for the analyses (and conclusions) of large amounts of data quickly by using the human experiences collected via technology. It provides a platform for standardization of the most efficient, effective processes. Much of the advancement in this area currently is occurring on the clinical side, where large amounts of clinical results can be processed quickly to provide physicians with the highest probability for the most effective treatment, whereas previous treatment decisions depended largely on personal experience and advanced training. On the financial side, large amounts of transactional information will provide foresight regarding revenue and cost performance, as predictors (volume, payer, type of service, historical payment patterns, usage rates, demographics, etc.) will enable healthcare executives to project revenue more accurately and address costs proactively.
In 20 years, wearable and implantable devices will predict diseases accurately, allowing for intervention before catastrophic issues. On the financial side, a blockchain-like transaction platform will be in place to captures all reimbursable (or value based-oriented) treatments, and push revenue to providers that are using artificial intelligence to automate the billing process (i.e. without human touch).
John Wood, PhD. CEO of Cardinal Point Healthcare Solutions (San Diego): "While precisely knowing what changes AI can and will bring to healthcare is impossible, it is certain healthcare will change. With that in mind, the best healthcare systems can do to prepare is more effectively manage external change and implement internal change. Mastering change management processes, training physicians to lead change initiatives, and developing fundamental communication strategies that keep all parties informed of what changes are occurring, when they will occur and why they are necessary, will set healthcare systems up for success regardless of the path AI takes in healthcare."
Ashish Koul. Senior Vice President and General Manager for Servion: "Artificial intelligence is expected to revolutionize healthcare in the near future. AI can have a positive effect on:
1) Healthcare customer service: Customer service bots in the future will be able to comb through one's medical records, issue and renew prescriptions, schedule appointments and even help with billing and administrative needs. This will lead to reduced human resources costs (due to fewer call center agents needed), faster issue resolution for patients and a more fluid overall customer experience.
2) Medical diagnosis: AI in the future will mine medical records, spot trends and even detect abnormalities faster. First steps are being taken in oncology and genomics today. AI will also be able to gather and analyze data from phones and smart devices. This can help improve the quality and speed of medical diagnosis.
3) Reach and availability: Technologies like natural language processing will allow virtual bots to converse with any individual having a cell phone. This is especially important in areas with less established healthcare systems, potentially providing basic healthcare access and guidance to a large number of people."
Suvas Vajracharya, PhD. Founder and CEO of Lightning Bolt Solutions (San Francisco): "There is real movement in the development of AI technology that makes physicians more efficient and safer.
At my company for example, our scheduling tool uses AI to drive greater operational efficiency, allowing patients to be seen earlier while avoiding physician burnout. There are also tools helping physicians more directly such as using AI technology to help diagnose and recognize patterns in X-ray films that may not be possible for a radiologist to see with a naked eye. Another example is the digital storage of medical records, which in itself is a very small step, but large amounts of medical records in digital form leads to big data which contain training material for an intelligent system that can suggest diagnoses.
In healthcare, physician burnout is increasing and the physician shortage is worsening, greatly impacting patient outcomes as access to treatment and clinicians becomes an issue. As more people recognize that AI can serve as a solution we can expect to see more widespread adoption."
Amy Simpson. Director and Industry Principle of Healthcare for Pegasystems (Cambridge, Mass.): "We are starting to see experimentation with AI in the behavior process area, particularly with managing health data and connected devices. As the amount of patient generated health data from wearables and other connected health devices continues to increase, AI and machine learning will enable care providers to mine patient data to make more informed and timely decisions to personalize care and improve health outcomes. For example, AI can detect smaller but identifiable changes in patient behavior that could negatively impact their health (such as a drop in glucose levels in a diabetic patient) and alert care management teams to preemptively intervene and ensure adherence to personal care plans.
The scale of this evidence base will dramatically inform the identification, treatment and management of chronic conditions. For example, changes in sleep, heart and breathing rate patterns acquired through devices may be matched with genomics and other biometric data to better predict which individuals will be Type I or II diabetics, and when. Fall patterns, identified by home monitoring devices, correlated with clinical population data, will drive the timing and type of non-medical preventive interventions that today are subjective. Drug research and development will be better informed by ready, real-time access to trial participant data, potentially preventing or reducing the incidence of adverse drug effects. Chronic conditions will be further differentiated by type and subtype based on this vast data set."
Mike Hilton. Chief Product Officer of Accolade (Seattle, Wash.): "AI is a vital technology in healthcare and will be a foundational aspect of tech platforms over the next 10 years. AI will have a big role in eliminating fragmentation and complexity that exists today by streamlining the way information is delivered, shared, analyzed, and applied across the healthcare continuum.
There are massive inefficiencies – and avoidable costs – associated with getting people to the right care, at the right time, and in the right clinical setting. Over the next decade or more, we'll actually look back on AI as a transformative technology that was key to enabling more personalization around the patient, and efficiencies that enable more relevant and effective care plans. AI has the power to leverage data, interactions and science in ways that directly benefit patients, providers, health plans, health system, and employers, all of whom are looking for stronger efficiencies, health outcomes, and cost reductions.
The areas where AI has been blocked from being fully usable in the past, are areas where Accolade has focused extensively: integrating the right data sets into the platform, including member interaction data, and applying a science layer that makes that data truly usable in healthcare. On the data integration side, it’s important to have access to data in a consolidated way – such as rich data from providers, health plans, employers, and rich contextual data from member interactions. And on top of this rich data is a strong science and machine learning layer, which enable AI to work.
For providers, the AI benefits span personalization, predictive care, and efficiency. Personalization gives providers a far more relevant view into everything that’s going on with the patient – symptoms, issues, medications, life factors, behavioral health concerns, access to care, biometric data, etc. New algorithms can give providers the next best recommendation to maximize patient health, while increasing efficiencies, as they’ll have the insights and the power to do things that were tough to do in the past.
We’re seeing the beginnings of the AI impact in our platform in the way we integrate and present data to our Accolade Health Assistants – who are on the front lines with our members working through care- and benefits-related concerns. Our Maya intelligence engine, with AI and machine learning at the core, crunches hundreds of data points every day, learns from the personal interactions with our health assistants and clinicians, and gets smarter, more personalized, and more accurate over time. AI gives massive stores of data an effective path to improving healthcare.
For the healthcare industry to continue to progress in delivering quality and affordable care, AI should be addressed, embraced, and implemented."
Mark Rolston. Founder and Chief Creative Officer of argodesign (Portola Valley, Calif.): "I look forward to a day when my relationship with my doctor is no longer episodic but instead continuous, supported by sensors that track all of my vitals constantly. AI cognitive systems are constantly monitoring this data and flagging conditions that my doctor needs to see. In this future my doctor knows me well. She will also have the statistical and cognitive support of the millions of other people tracking their own health. In this future, information is shared not just though papers and conferences but though a cognitive system that is constantly processing through all of our collective data to find 'the needles in the haystack' giving doctors amazing new levels of diagnostic insight."