As part of a broader transformation of the health value chain, AI and machine learning promise to help change the role and business models of life sciences – by accelerating scientific breakthroughs, identifying previously elusive patterns in unwieldy global data masses, and enabling greater drug personalization.
Here, Siniša Belina of AMPLEXOR Life Sciences explores some of the potential opportunities - and how to prepare for them
Artificial intelligence is already making its mark in many everyday situations, at home and in the workplace, and is something companies in all industry sectors to start planning for – including life sciences. Automated personal assistants such as Alexa and Siri, smart web or content searches that adapt to user preferences, and customer care channels such as web chat facilities are already exploiting AI in everyday situations. Via machine learning, a subset of artificial intelligence, algorithms don’t just make clever connections and spot trends in masses of data; they also become increasingly refined and efficient at this over time, in response to the conditions they are exposed to and the results they find. All of which adds to the speed of discovery, and the next actions this makes possible.
Automation is a big attraction of the AI proposition. If machines can get to grips with routine knowledge work, and do it more rapidly without needing breaks to sleep, rest and refuel, then it makes sense to apply technology to sift, fill, find and organise. As long as humans are overseeing and sense-checking the results, why not let IT systems take the load and let experts do the more interesting and mentally demanding tasks?
Smarter health
In primary care, AI-based systems are already beginning to play a more direct role in patient diagnostics. UK researchers at a hospital in Oxford recently announced the availability of AI technology that can diagnose heart disease and lung cancer at a much earlier stage from analysis of patient scans . The heart disease technology is due to be made available to NHS hospitals for free sometime very soon. It makes sense that overstretched health services should take advantage of the technology capabilities open to them, especially now that high-capacity computing resources is readily accessible via the cloud, and given the mounting pressure on staff.
We are likely to see smart robots help with the heavy lifting of patients, as they are moved in and out of hospital beds, saving the backs of nurses and support staff. Connected devices, meanwhile, will be used increasingly to relay patient data to the health practitioners managing their care, not only to monitor the progression of or any improvement in their condition but to allow much earlier interventions if the continuous trackers begin to pick up signals that indicate certain types of subtle changes.
Ongoing patient monitoring is part of a strategy for more pre-emptive, preventative care – a shift towards maintaining wellbeing rather than reacting to illness. All of which is expected to lead to better outcomes for patients, and a reduced strain on healthcare resources, hospital beds, and so on.
Want more? The Future Of Drug Discovery: AI 2020
Pharmaceutical industry innovation
So where does this leave big pharma and the wider life sciences industry - which has done so well, for so long, from providing treatments that alleviate, heal and manage existing patient conditions? How far might AI take them in transforming the way they operate, and indeed the role they play in the health cycle?
The surge in technology-related events for the healthcare and life sciences industries is no coincidence. This year, conferences and exhibitions are seeking to bring new awareness of the opportunities to the combined sector. In April, 'Next Generation Healthcare' formed the central theme of theme at BioTrinity 2018 , whose sessions include 'AI and Drug Discovery'. And this summer, Artificial Intelligence summits are being hosted everywhere from London to Philadelphia . Alan Boehme, CIO/CTO of Procter & Gamble, and Juliet Bower, Chief Digital Officer at NHS England, featured among the speakers at the recent London event. Even if it’s just to have a response ready for funders, partners and patients, industry leaders recognise they need to have a position on AI.
An industry event in June highlighted the technology’s substantial potential for enhancing research and development operations, through the ability to analyse large volumes of data leading to richer insights. To this end, applications, systems, and platforms have already been developed to transform clinical trial innovation. This isn’t just about teasing out finer details and subtler patterns from once untameable volumes of disparate data either. It’s also about modelling and extrapolating from such findings to arrive at bolder hypotheses and deeper and more targeted work, accelerating progress.
Beyond traditional drug development, AI and machine learning in particular offer scope for new advances in medical imaging interpretation, genomic profiling, personalised medicine and treatments.
Reconnecting with patients
AI technology also offers a way to track global patient trends, concerns, experiences, behaviour and needs, enabling the life sciences industry to understand what is happening in the real world to a level of granularity and completeness that hasn’t been possible before. This offers potential not only for more proactive and thorough monitoring of adverse events and other safety signals as drugs move into markets, but also for spotting untapped requirements, triggering new innovation.
Where the life sciences industry has traditionally been one step removed from patients, public internet forums and social networks offer an opportunity to understand evolving demands and engage with patients in new ways. Although companies have to be very careful about disguising promotions as neutral information and advice, a greater dialogue with patient communities could be their best shot at capturing a share of the growing wellness/preventative medicine opportunity. The global nutraceutical market, valued at around $383.06 billion in 2016, is expected to be worth $561.38 billion by 2022 , such is the growing consumer appetite for products that keep them healthy. Other research has shown that Millennials are prime targets for proactive treatments in this category – a demographic that is very vocal on social media.
Overcoming inertia
The scope for AI in transforming life sciences as we know it today is great. But this is not a fast-moving industry, and there are a number of things that need to happen first if companies are to adapt to and exploit the potential ahead of them in a sufficiently timely fashion.
The first is a recognition and acceptance of the fact that change is coming, and that no industry is immune to disruption from emerging market entrants – new potential competitors with bold ideas and the advantage of not being tethered to legacy thinking and ways of working.
The second is preparing an IT and data environment that allows for new experimentation and insights – within the restrictions of regulatory control and privacy protection, of course.
Already, today, the world is building knowledge at an unprecedented rate: IBM estimates that, by 2020, knowledge will double every 11-12 hours (compared with a rate of every 25 years, as was the case in 1945) . So there is a growing urgency for companies to bring this situation under control in their own context, and harness it to maximum potential.
This isn’t just about developing ‘big data’ strategies, but organising and preparing that data so it can be analysed efficiently, accurately and holistically using AI platforms – to spot emerging trends, anomalies, concerns and opportunities, at a speed and degree of precision which in time could become a market differentiator.
Starting from safe territory
For now, regulatory pressures are behind a lot of data-related initiatives in life sciences. More data is being captured, consolidated and cleaned up now, but primarily this is for a specific purpose, and not one that will add significant value for the business. Innovation is not a part of the plan, when it really should be. If the work has to be done, far better to do it once and do it well – laying the foundations for all manner of future use cases, however futuristic these might seem now.
And, as mundane and mandatory as regulatory data initiatives might seem, they do in themselves offer a potential platform for experimenting with AI. Using machine learning, for instance, systems could ‘learn’ how to produce better output, or the conditions most likely to result in a new marketing submission being accepted first time.
The critical enabler for all of this is the creation of a comprehensive master data model – one that also includes inter-dependencies between the data, so sources can be exploited to maximum potential. Beyond that, companies need a strong sense of new purpose and direction to bring all of the potential to fruition. The best innovation starts with a big idea.