Tech is the most infectious virus the world has ever seen. Industries are being “attacked” on a daily basis from technological advancements with different levels of success, adaptability and impact. Mobile was one of the first industries to catch fire from the rise of tech with mobiles becoming nothing like their predecessors. Healthcare seems to be next in line.
Remember flip phones and their uncanny ability to…call people? Mobiles have now turned into entertainment units, TVs, working stations, credit cards, music players, fitness trackers and so many more things it would take us days to list. Technology didn’t fix phones, it completely redefined them.
The same logic applies to the relationship between the term artificial intelligence and healthcare. The industry’s interaction with strong AI has been tentative at best, with impact and results only scratching the surface of what could be achieved. A 2018 Accenture report puts the potential of AI in healthcare into perspective:
“Growth in the AI health market is expected to reach $6.6 billion by 2021—that’s a compound annual growth rate of 40 percent. In just the next five years, the health AI market will grow more than 10×2.”
Moreover, as stated by The State of AI: Artificial Intelligence in Business, 62% of businesses are thinking of investing in AI soon and 42% think that AI will be important for their organisation or industry in the next 5 to 10 years.
Predicting, preventing and curing disease will look nothing like it does today in a few years’ time. Machine learning and artificial intelligence (AI) are slowly yet steadily infiltrating the industry on so many different levels and the potential of what can be achieved is truly astonishing.
The building blocks that healthcare is currently built on will most definitely be replaced by new, innovative ways of doing things more efficiently. Just think of the basic health check up. People have to visit a hospital to get a general test on their blood pressure, heart rate, fat levels etc.
Can you imagine a world where a wearable watch with built-in ECG features can monitor your health all year-round? The health paradigm changes entirely as everything has to change: doctors have more time, patients are more informed and emergency incidents will be few and far between.
Data, Data, Data
Get this, there’s more health information outside the health system and health records than inside it. Marketing departments of big grocery chains hold so much information on our eating habits and our nutrition. The Internet is full of knowledge on the advancements of healthcare and Fitbit activity trackers run checks on your health on a daily basis.
This information is scattered all over the place and the sources are random and diverse. If you could combine the enormous amount of data sets found outside the healthcare system and combine it with EHRs, the sky ‘s the limit for what AI can do.
If AI is a driving car, data is the fuel. The more data you have, the more AI development can flourish and work its magic. The benefits of gearing our healthcare system towards a data-driven model are countless. Data extraction leads to insights and insights lead to outcomes.
Think of the pharma industry. It provides products whose success is not necessarily judged on results. Medicines and painkillers work like every other industry where demand and supply determine prices and product movement.
What if data changed the way this whole process worked? What if the pharma industry worked on data-driven performance indicators and results? Companies would then have to try more on their product quality, research and development and overall performance, knowing they will only get paid if the product delivers.
The Biotech industry and drug development are two of the most important sectors that will be affected by AI systems. The business moves by Big Pharma companies are proof of the faith these conglomerates are showing in AI, investing money and resources in the acquisition/collaboration of AI startups.
Roche has partnered with Owkin and bought Flatiron Health, whilst Novartis has joined forces with Massachusetts Institute of Technology, IBM Watson, Quantumblack, and Intel to propel its AI ambitions. Moves like these solidify the commitment of the industry to AI and what it can do for Pharma as a whole.
Let’s dig a bit deeper.
Did you know that 9 out 10 clinical drugs fail to make it to trials, and a lot more don’t reach [FDA] approval stage? What does that mean for Big Pharma companies? Costs of drug discovery and development are insanely high. They put all their eggs in that 10%, hoping that the one drug can cover the costs of the failed attempts and make them some profit on top of that.
Total costs surrounding development and approval of new drugs compound to a staggering $1.2 billion according to a 2016 report by Elsevier. These costs are eventually passed down to the customer, making the finished product unattainable for a big chunk of the population. The goal with AI is to create cheaper, accessible healthcare for everyone. Cutting costs for Pharma companies along the way is a byproduct of that.
The signs are more than positive as AI and deep learning are already accelerating drug discovery for the industry’s biggest players. Just look at Novartis and how machine learning algorithms are changing the way they do science.
This interesting article gives us access to behind the scenes insights on how Novartis is implementing machine learning in action during drug discovery. It is staggering to see that AI is no longer the next big thing in pharma but it’s actually the current big thing. The level of adoption is growing rapidly and we should expect to reap the benefits very soon.
The use of AI and machine learning is opening new horizons for scientists, allowing them to see medicine from a scope and lens that only advanced tech can. This quote by Jeremy Jenkins, Head of Informatics for Chemical Biology and Therapeutics at the Novartis Institutes for BioMedical Research (NIBR) says it all:
“Machine learning is pointing us to new therapeutic possibilities with unprecedented efficiency…it has an unparalleled ability to teach us about how our drugs are working.”
We’re already at a point where AI has matched doctors when it comes to diagnosing a condition or a disease looking at images. According to AI research led by the University of Birmingham and University Hospitals Birmingham NHS Foundation Trust, AI is at this point capable to match the diagnostic abilities of human doctors, identifying anything from an eye infection to more serious diseases like cancer.
Even though the sample size of the research is not big enough to allow AI to step into the role of a doctor, it is sufficient evidence for optimism about the potential of AI in the healthcare space. The findings open the floodgates for a discussion that can revolutionize the medical diagnosis procedure as we know it.
You’re at home and you have a bad cough. You pick up your phone, you open an app and you couch. The app picks up and interprets the sound, accurately diagnosing the lung problem at hand, prescribing you the right syrop.
Self-diagnosis tech is where AI can make heads turn and bring real value to the industry. Apps like Ada, Babylon and SkinVision are just the beginning of what promises to be a bright future for healthcare diagnostics. These advancements will be huge for complex problem solving, freeing-up resources, speeding up the diagnosis and letting the doctors deal with curing patients.
Having your own personal assistant in a time of need is not as futuristic as it sounds. Virtual nurse assistant company Sensely was able to raise $8 million in Series B funding whereas Care Angel’s virtual nurse assistant Angel is another great example of where the virtual nursing assistant industry is at.
A Syneos Health Communications report revealed that 64% of patients admitted they would be comfortable with AI virtual nurse assistants, emphasizing on benefits such as 24/7 access to answers and support, continuous monitoring, and swift responses on medication questions.
Predictive Analytics in Healthcare
A Forrester Consulting survey raised the point that a measly 34% of healthcare organizations have embraced AI-powered predictive analytics, which comes in sharp contrast with the 51% in other industries. The slow adoption is showing promise of an increase as the development is both rapid and accurate.
The use cases for predictive analytics come in many ways, shapes and forms. Remote health monitoring is one of the best examples to showcase the AI prowess. Just look at companies like Alignment Healthcare and how they are reshaping the predictive analytics space.
Their virtual application command center is a “take home medical kit” that allows people to get basic health information such as blood pressure and weight. The AI-powered platform runs the findings through the system and is able to identify and report on potential health risks and advice on the desirable cause of action. That could be advice on medication or visiting an onsite doctor.
This remote monitoring program can come in hand for older people that usually need a home-nurse to check up on them in frequent intervals.
Then you have chronic disease. The silos of data on chronic disease is fuel to the AI fire. Give the algorithms tons and tons of data sets and use cases on diabetes, asthma, arthritis and Alzheimer’s disease and what you have is a predictive analytics wonderland with prognostications that can improve survival rates and make life easier for people that suffer from these diseases.
Healthcare Administration & Workflow
This is the area where AI has already been used the most. Digitizing long and arduous administrative processes can elevate the level of customer service, customer satisfaction and overall healthcare experience. Using AI in conjunction with computer systems can solve problems like scheduling, appointments, medicine/vaccine stock count, patient registration processes and so many other admin areas.
AI can perform tasks that will free-up healthcare professionals to deal with more important aspects of the healthcare process, taking over tasks that can be optimized and automated. AI solutions provider Nuance is a good example of how computer-assisted physician documentation (CAPD) can have an immediate impact by providing accurate clinical history and consistent recommendations to patients.
Does AI Replace the Doctor? Ethics & Decisions
The case in favour of AI taking center stage in healthcare is getting stronger and stronger. Just like every other sector that has been invaded by tech and AI, the question of ethics and whether the machine will end up replacing humans, is too obvious to avoid. More on this subject on our recent article “Debunking Tech Myths: Technology Is Taking People’s Jobs”.
Could we foresee a scenario where machines are projecting, diagnosing and curing patients without any human involvement? No. Even though the saying “never say never” should be the tagline for AI technology due to its unlimited potential, healthcare is a unique sector that guarantees the involvement of human doctors despite the advancements made in AI.
The human touch and the fact that humans know the feeling of pain, a cough, a headache and the entire experience of not feeling well, will always come in handy in medicine.
A recent Deloitte article on “Predictive analytics in healthcare” raises the question of accountability between man and machine and the risk related to making crucial decisions. Who bears the last word and who is responsible for mistakes or even success? Just read the following passage from the article:
“The transfer of risk and liability within the medical industry is complex and this risk combined with misdiagnosis from a machine adds to the complexity that needs to be addressed when integrating predictive analytics into health care.
This could increase risk in health care if, for example, a doctor relies on a computer to give a diagnosis over their own assessment. They may take more risks because they believe they are protected with the computer being accountable and bearing the cost of the risks. This challenges the ethics of respect and doing no harm, with the key decisions being outsourced to a machine and the accountability lines being blurred in the diagnosis and treatment plan.”
The lines are indeed blurry and balance of power is an interesting area of discussion. Implementing AI in healthcare is not just about results, performance and use cases but it has a lot to do with the relationship, roles and jurisdiction of man and machine. This is something that will keep developing in tandem with AI and it’s an area worth keeping tabs on.
Healthcare is its own little universe and it comes as no surprise that it will take some time for AI to fully interact with every single part of this ecosystem. The level of success and maturity of product differs from sector to sector but the one thing that is consistent throughout is this: AI is the present and future of healthcare.
The building blocks are in place and the foundation of the industry are being reshuffled and reshaped on a daily basis. Watch this space as we will be checking back at this interesting relationship between AI and healthcare, monitoring new advancements and updates.