In 2013, US-based startup Berg was requested by the Department of Defense to assist in improving the detection of prostate cancer, which is an illness that is common amongst pilots.
About five years later, an AI-driven tool has been tested on over 1,000 patients, and it has shown promising outcomes.
The excitement surrounding big data, machine learning and artificial intelligence have triggered increased growth in health-technology startups, particularly in a market that was traditionally controlled by big pharma.
In spite of the debate over the degree to which artificial intelligence will change medical science, billions of money are currently being bet on a groundbreaking thing coming up to propel the next innovations in drug development.
This year (2018) has seen a boom in investment from big pharmaceutical companies and a range of partnerships with health technology organizations.
According to an analysis by Deep Knowledge Analytics, there are about 15 companies integrating artificial intelligence (AI) into their drug discovery processes in 2018.
“We’ve seen huge interest from many of the major pharma companies” in the last 18 months, said Andrew Hopkins, the chief executive officer of Exscientia, an AI-powered drug discovery firm working with GlaxoSmithKline.
Other collaborations include Sanofi’s and AstraZeneca’s initiatives with Berg, as well as Merck’s project with Numerate, a drug design company.
In-house projects include Pfizer utilizing artificial intelligence in mining patient data in a bid to identify signs of an unusual heart failure condition.
MORE – Artificial Intelligence in Medicine – Top 10 Applications
MORE – Data Science – 8 Powerful Applications
Novartis looks forward to a drug part-created with artificial intelligence (AI) being registered in the next three years.
“We see ourselves shifting from the mindset of a traditional pharma company to [one] more inherently flexible, typical of a tech company,” said Badhri Srinivasan, Novartis’s head of global development operations.
According to consultancy IP Pragmatics, nearly $5 billion was invested in artificial intelligence firms in 2016, with healthcare as one of the rapidly growing sectors.
By 2021, the value of this sector is anticipated to hit $6.6 billion, with a significant portion of the growth in China.
In the United Kingdom, about five new government-financed technology centers are expected to be opened in 2019, by utilizing AI to expedite disease diagnosis with the goal of increasing the efficiency of the National Health Service.
Startups in the early-stage drug discovery space including BenevolentAI utilize algorithms in sifting through massive volumes of data to find patterns that humans alone may not identify and produce various hypotheses to medical issues.
Hypotheses yield possible solutions: firms like Insilico are currently utilizing AI in designing treatments that have not been discovered in chemical libraries or nature; others utilize artificial intelligence in stimulating clinical trials, prior to the selection of real-life candidates.
In a sector that is currently experiencing miserable statistics, hunting for efficiencies is vital.
It takes about 12 years and a cost of $12 billion to make a drug available in the market.
Worst case scenario, most of the clinical trials do not bear fruit.
According to Deloitte’s report, the forecasted returns, particularly on drug research and development investment dropped to their lowest levels in 2017, whereas spending grew.
According to Berg’s CEO Niven Narain, inaccuracies and inefficiencies mean “all we’re doing is magnifying the trial and error process.”
Costly drug prices are owed to increased attrition rates, as pharmaceutical firms look forward to offsetting the cost of unsuccessful projects against the few that succeed.
Boosting productivity “hopefully will have downstream effects,” said Julie Schiffman, Pfizer’s VP of business analytics.
“AI won’t necessarily drive the cost of drugs down,” says Eric Sandor, of consultancy firm Genpact. “What it will probably do is identify drugs that can be specifically formulated for small groups of patients . . . that are more specialized, which could, therefore, be more expensive.”
According to Simon Smith, the chief growth officer at AI-powered biomedical search engine BenchSci, artificial intelligence (AI) does not work like a silver bullet. “The risk inherent in developing a drug isn’t going to go away just because you used a machine to do it,” he says.
Deep Knowledge Analytics stated that in spite of the latest hype surrounding artificial intelligence technology, there appears to be “promising” drug discovery startups and inadequate artificial intelligence experts.
According to Mr Smith, moving past the “low-hanging fruit” to innovations that make drug development cheaper or faster would be “much harder” compared to achievement.
Mr Smith also said that the existing algorithms could be concentrating on the wrong things. “If the processes themselves aren’t going to make the next big breakthrough, it doesn’t matter if we expedite them by 100 times.”
