Artificial intelligence is making its way into healthcare with clinical applications such as diagnostics support as well as operational efficiency applications to upgrade practice workflows and supply chain management.
Twenty years from now, AI could make healthcare more precise and deliver better outcomes; however, it's unlikely machines will completely replace physicians and nurses at the point of care. Here, 13 medtech leaders answering the question: What role will artificial intelligence play in healthcare over the next 20 to 30 years?
Waqaas Al-Siddiq. Founder, Chairman and CEO of Biotricity (Redwood City, Calif.): "The capability of AI to improve over time as it amasses more data and learns and refines its algorithms is well-suited for medical device technology. AI technologies have the potential to produce an even greater effect within the healthcare industry by reaching patients directly via mobile devices. Initially, doing so will center on connectivity issues and the need to offload data via the cloud, as most endpoint devices in the internet of things do. Developments in computer chip processing speed will result in a situation in which AI functionality is embedded within the devices themselves, profoundly changing the way medical devices work. The impact on the healthcare system will be a change in practice workflows, care delivery and the integration of devices that make treatment recommendations as opposed to simply providing data for physician consumption.
As AI and machine learning are further integrated with chronic care management and remote monitoring, we will see the current physician practice patterns, which have become focused on gate-keeper and clerical functions, disappear. These technologies will empower physicians by allowing them to focus on those areas that truly require their diagnostic and therapeutic skills, including physical intervention such as surgeries, bone adjustments in the case of broken bones, individualized cancer treatments, unique outlier cases and how even routine cast that does not require a bone adjustment can be performed with a 3-D printed device.
No doubt, deep learning and machine learning are notable as a tangible demonstration of artificial intelligence ability to improve costly chronic healthcare management and a means of implementation through mobile technology-based wearables. This idea is central to the mainstream adoption of artificial intelligence, particularly when one considers the abundance of digital assistants in contemporary smartphones and the obvious mobile implications of autonomous vehicles. By aligning itself with the broadening mobile capacities of contemporary society, artificial intelligence can become as influential, and mainstream, as mobile itself."
Paul Clark. Director of Healthcare Research for Digital Reasoning (Nashville, Tenn.): "In 20 to 30 years, artificial intelligence will be a standard component of daily operations for hospitals and physicians — invisibly augmenting and amplifying human effort — with as little fanfare as an email or X-ray. Over the next few decades, AI will automate the healthcare tasks that require reading large volumes of unstructured data and making decisions.
Healthcare providers should be preparing now by starting with developing a comprehensive data strategy, beginning with the underlying data quality, acquisition, integration and disposition. After creating an effective data governance and big data architectural framework, providers can move into creating organizational data science expertise and embedding big data analytics tools which combine to produce exponential return-on-investment results in care efficiency and efficacy.
The only way healthcare providers can keep up with the coming multi-decade AI wave is with a foundation of a high-quality data framework, data science expertise and world-class unstructured big data analytics tools."
Leonard D'Avolio, PhD. Founder and CEO of Cyft (Cambridge, Mass.): "AI is already seeing success first in the business of healthcare in billing, scheduling and logistics planning. After gaining the trust of C-level folks in terms of it being 'real' and delivering return-on-investment, it will move into clinical areas that involve assessing large volumes of quantifiable data to make assessments. Expect pathology and radiology to be fundamentally transformed in the next five years. Diagnostics that also meet the criteria of producing and relying on quantifiable data will also change in that time. For example, tests that rely on imaging or chemical reactions will become more accurate as a result of assessing larger volumes of outcomes data. For that reason, diagnostics companies will form relationships or [develop] products that bring outcomes data back to them in some form.
The trend will not lead to AI as a doctor so much as AI-enabled doctoring. For the first examples of AI behaving doctor-like, look to large academic medical centers. To see it transform healthcare, look to the more resource-constrained settings. Their incentives are aligned to give PCPs greater access to specialty expertise. They are already ahead in remote prescribing, automated medication delivery systems, telemedicine, etc."
Manan Goel. Vice President of Product Marketing for Kinetica (San Francisco): "With the nexus of people, devices, data and AI, the internet of things will have a profound impact on every aspect of [the] healthcare ecosystem. Healthcare organizations will leverage interconnection of devices, systems, processes, facilities and equipment with AI to collect, exchange, interpret and act on information. AI-driven healthcare IoT will streamline processes such as supply chain, inventory management, equipment maintenance, payments and drug development to reduce latency, lower cost and deliver operational efficiencies.
Identifying and eliminating fraud, waste and abuse requires self-learning systems that can crunch through massive volumes of complex data to discover hidden patterns, anomalies and relationships. AI can help build such systems, quickly uncover nefarious activities and suggest corrective actions to minimize fraud, waste and abuse.
Invest in your people so they pick up additional AI-related skills. Look for non-proprietary, industry standard solutions so it's easier to find the people and skills to manage the AI systems. Establish business process to balance AI-driven innovation, automation and mechanization with governance, control, traceability and auditability to build trust and acceptance for AI-driven automation."
Karin Lachmi, PhD. Chief Scientific Officer of Bioz (Palo Alto, Calif.): "Adopting artificial intelligence and emerging technologies as a whole will be crucial in order to accelerate health and life science research and ultimately save lives. Not only does this include embracing AI, but also finding better ways to reach researchers, scientists and MDs through more efficient, transparent, structured and aggregated knowledge of scientific data.
In this spirit, over the next several years journal publishers will experience increased public and government pressure to make more scientific articles publicly available — especially with the decline of printed journals in favor of online article delivery. There are hundreds of millions of pages of scientific papers, and AI can help researchers and MDs sort through the findings in order to better plan for experiments, speed up the drug discovery process and find cures for diseases."
Neil Patel. President and COO of Healthbox (Chicago): "A lot has been said about the potential impact of artificial intelligence and machine learning in healthcare. These predictions range from replacing part of the physician workforce completely to enabling all physicians to be more efficient.
In the near term, AI has the potential to impact radiology and pathology. These advancements will provide physicians with greater insight and increase the speed in which they can make accurate diagnoses — effectively giving them superpowers. In the long term, we can see some basic functions of radiologists and pathologists being replaced completely by AI and leaving the more complex diagnoses to human physicians.
AI needs massive amounts of data to learn and high quality data to work in real time. Health systems can prepare by making sure their data architecture, security and storage are sound and set up for the ever-increasing amounts of data being generated and analyzed. Physicians can prepare by embracing and training for AI rather than fearing it."
Michael Sierra, MD. Vice President of LEO Science & Tech Hub (Cambridge, Mass.): "There is a genetic predisposition to disease but that is not always enough. Environmental factors also play a role in the onset of disease. Two identical twins, genetically equivalent, separated and living in two different regions of the world, one develops disease the other is free of disease. Our smartphones have a GPS that captures temperature and humidity, as well as apps that can track diet and exercise. All of this data can be collected and with AI/deep learning identify the relationships in the data to predict disease. So, like the weather, having a 50 percent chance of rain, you are informed that you have a 50 percent chance of a flare-up in the next two weeks. You are able to predict disease and treat to prevent rather than only treating the symptoms."
Gurjeet Singh. Co-founder and Executive Chairman of Ayasdi (Menlo Park, Calif.): "The power of AI to find patterns in all of this unimaginable data, and use, for instance, an individual's genomic structure, to tell us what treatment would work best for that person; this will be the norm in an AI-driven healthcare system.
AI will predict patient health with high accuracy. AI will power predictive models that will give healthcare organizations foresight into the future needs, costs, disease burden and risks of patients. Determining the patient groups that are projected to have the highest escalation of costs over time, the conditions likely to appear for each group, and an individual's predicted change in utilization will all become the norm. Predictions will be across multiple targets, taking into account health- and non-health-related information to determine an individual's likely course. And it will do this in a completely transparent way, providing reasoning behind its assertions and augmenting human skill and knowledge. By providing a thorough understanding of the 'why' behind predictions, AI will make its way into day-to-day healthcare decision-making.
AI will advance precision medicine. The promise of precision medicine relies on tapping into all available data about an individual. AI's capability in unsupervised discovery will help us find subtle nuances and patterns in huge datasets of medical records and genetic information. Breakthrough discoveries in mutations and linkages to diseases, predicting what may happened when DNA is altered by genetic variation, and learning of new disease genes will allows us to execute on very personalized treatments and interventions and prevent an array of diseases and illnesses."
Kim Terca. Director of User Experience and Press Relations at Netvibes (Paris, France): "By connecting data, devices and people together, the internet of things is poised to change the healthcare industry, especially as artificial intelligence continues to grow, enabling software to do more things with data. Some of this is already happening today. For example, consumers can now program their smart scale to automatically record all weight measurements in a cloud spreadsheet, monitoring their weight goals over time, or even share the data with their doctor. All of this can be managed online from a central dashboard, aggregating data together and allowing users to program their own AI actions. In the future, healthcare responses could be fully automated. Imagine Joe's smart watch detects a heart attack. Based on Joe's location and customized settings, the watch could automatically call for an ambulance and send a text alert to loved ones."
Tryggvi Thorgeirsson, MD. Co-Founder and CEO of SidekickHealth (Palo Alto, Calif.): "AI will bring about radical changes to every aspect of healthcare, including patient outcomes, workforce requirements and operations. It will become a defining competitive factor across the whole spectrum of healthcare provision. The impact of AI on healthcare will be no less than the effect of automation on manufacturing.
Staff will gradually transition from working in parallel with AI such as with current clinical decision support, to being more and more hands-off and focusing on edge cases and oversight. Amongst other things, this will allow caregivers to spend more time bedside with their patients; something that many currently struggle to find time for.
It will be interesting to see whether the resulting increased efficiencies will be able to keep up with or even outpace the ever-increasing needs associated with the increasing longevity that these changes will help bring about."
Matt Zeiler. Founder and CEO of Clarifai (New York City): "Artificial intelligence is already having a hugely positive impact on the healthcare industry — from detecting cancer and treating skin problems more accurately, to helping deliver reliable and cost-efficient medical devices to physicians all over the world. In 20 to 30 years, these tools will only get more accurate as they're fed more data and used more frequently. In 20 to 30 years, these tools will be the de facto standard in medicine as their accuracy is already exceeding that of humans in many different industries, including medicine. The next phase will not only involve continued improvements in AI algorithms and data collection, but distribution of devices that gets this technology outside the hospital to patients all around the world."
Srinivas Kowta. Senior Director of the Health Analytics Vertical at Axtria (Berkeley Heights, N.Y.). "AI will revolutionize medicine by improving the accuracy of the decision-making systems in identifying treatment options in regular and emergency setting. It will also accelerate the time to diagnose hard-to-diagnose conditions such as cancers, Alzheimer's, multiple sclerosis and other rare disorders.
AI has the potential to increase the alternatives in treatment choices when the primary alternatives may not work in real time and customize healthcare to a segment of one. It could exponentially increase the efficiency of physician based repetitive tasks and accelerate new drug discovery."
Seth Krauss. Product and Architect at Kickdrum (Austin, Texas): "While there are many possible applications for AI in healthcare, currently, there is a heavy focus on automating tasks that involve analyzing sight and sound. What that means is that there is a big opportunity for AI to become a major player in a specialty like radiology, particularly with X-rays, assisting in the initial processing of images and differentiating what is a normal scan versus what is abnormal in a scan. It also means that patient charting is an obvious target for natural language processing. Dictation and NLP will bring increased standardization and reduce errors made due to an inability to correctly understand or read notes.
Despite all of the technological and AI advances to prepare for, physicians should be excited for what's to come over the next few decades. AI and automation will inevitably become more prevalent in healthcare, but fully automating many responsibilities doctors have will take big developments in tech that don't appear on the horizon quite yet. However, it does appear we're on pace to automate enough to allow doctors to better prioritize their patient workload, allowing for AI capabilities to replace what were previously time-consuming, less-critical tasks."