Category Archives: Artificial Intelligence

The third wave of AI in pharma R&D

Wax Selection – Leaders in Pharma, Biotech & MedTech Recruitment

The capabilities of artificial intelligence are advancing and its ‘third wave’ offers the ability to analyse enormous sets of data, identify patterns, and generate algorithms to explain them, to the benefit of researchers.

The digital revolution vastly accelerated the research, development, and production of new drugs. Digital technology has augmented the natural capabilities of researchers and scientists in a variety of ways. Now, artificial intelligence (AI) is poised to take this augmentation to the next level.

One of the most significant ways in which AI technology augments human capacity – particularly in an R&D context – is by automating repetitive, lower-level cognitive functions that once had to be carried out manually. This liberates drug researchers to focus on higher-level thinking. This advantage was identified early in the development of AI technology by J C R Licklider, who wrote back in 1960 in Man-Computer Symbiosis:

“About 85% of my ‘thinking’ time was spent getting into a position to think, to make a decision, to learn something I needed to know. Much more time went into finding or obtaining information than into digesting it.”… “Several hours of calculating were required to get the data into comparable form. When they were in comparable form, it took only a few seconds to determine what I needed to know.”

A related idea was expressed by Herbert A Simon with the concept of ‘bounded rationality. He wrote that humans’ decision making can be optimised when they are provided with a limited quantity of relevant, focused information and sufficient time to process it.

With the advent of contemporary AI technologies, the bounds of human rationality have been expanded. AI provides drug researchers with a greater breadth and depth of data that is simultaneously more focused and relevant than the data sets of the past, enabling researchers to optimise their decision-making capacities.

The continued advancement of AI will augment humans’ power of critical thinking in three key areas that are relevant to the medical and pharmaceutical industries: computing advanced mathematical problems, analyzing complex statistics, and generating novel hypotheses. These areas correspond to the ‘three waves’ of AI development throughout the 20th and 21st centuries.

The first and second waves

The first wave of AI development brought us ‘knowledge engineering’ optimisation programs, which solved real-world problems efficiently.

While applications specific to the pharmaceutical industry were scarce, the broader medical field benefits from first-wave AI technologies every day. Take the Framingham Risk Score Calculator, which utilises AI to predict the heart disease risk of any patient.

Machine learning programs were brought along by the second wave of AI. These solve complex pattern recognition problems using statistical analysis. Unlike their first-wave predecessors, second-wave AI programs perceive and learn – often as well as humans do.

Clinical decision support systems use second-wave pattern-recognition programs to analyse and evaluate medical test results. Similar machine-learning programs are beginning to be used by leading pharma firms in a variety of research and development contexts to predict drug effectiveness, to discover new compounds with pharmaceutical qualities, and to develop new combinations of existing drugs.

Second-wave AI is powerful, but it demands well-organised, consistently-coded, and complete data sets in order to accurately conduct its analyses. This limitation is now being overcome by the third wave of AI.

The third wave

We have now entered the third wave of AI development. Third-wave AI programs have the capacity to analyse enormous sets of data, identify patterns, and generate algorithms to explain them. These programs normalise the context of disparate data points and generate original, novel hypotheses at a faster rate and with greater accuracy than human researchers can.

Only in this third wave have AI programs reached a sufficiently advanced state to effectively analyse the vast and complex web of unstructured biological data. Until recently, biological data had to be manually cleaned and organised through extensive and costly human effort. Now, AI programs use a combination of machine learning, natural language processing, and text analytics to analyse unstructured data in real-time.

Through context normalisation, third-wave AI technology dramatically increases the quantity of data that can be analysed in the course of the drug discovery and testing process. Furthermore, it enables the simultaneous generation and testing of new hypotheses at a rate that would be impossible without such immense computing power.

Aided by this technology, drug researchers can arrive at a higher quantity of more accurate hypotheses and can test these hypotheses with unprecedented speed. The result is a significantly faster, and less expensive, discovery process, with lower risk and more effective results. Firms such as Pfizer and Johnson & Johnson are employing such methods to great effect.

Given that R&D consumes as much as 20% of pharma firms’ revenues, and that the total price of developing a new drug has ballooned by 250% in the last 30 years, it’s not surprising that firms are eager to embrace third-wave AI as a means of accelerating drug development.


AI predicts heart attack risk factors from retinal scans

Wax Selection – Leaders in Pharma, Biotech & MedTech Recruitment

Artificial Intelligence can be used to predict heart attack risk using retinal images, according to new research backed by Google.

The researchers trained deep learning algorithms on data from thousands of patients recorded in a massive UK study, which was used with retinal scans to produce a program that can identify risk factors from the scan information alone.

This predicted cardiovascular risk factors not previously thought to be quantifiable using retinal images: these included age, gender, smoking status, gender, systolic blood pressure and major adverse cardiac events.

Reporting results in the journal Nature Biomedical Engineering, the team used a dataset from the UK’s Biobank, a study where 500,000 were recruited between 2006 and 2010, and agreed to have certain health measurements recorded.

Health outcomes such as hospitalisation, mortality and cause of death were also logged. Smoking status was obtained via survey using a touchscreen interface in the research backed by Google and Verily, Alphabet’s healthcare subsidiary.

Participants were also asked to identify whether they are a current smoker, former smoker, or had never smoked, and blood pressure readings were also taken.

A further 67,725 had paired images of their retina fundus taken, along with a second group that was used to create a training dataset with known risk factors.

The researchers asked a neural network to make an output prediction based on the fundus image.

It was able to analyse images from the group with unknown risk factors, compare it with the training data set and after the process was repeated, was eventually able to predict cardiovascular risk factors from new images.

This is just the latest in a series of studies showing the potential of AI to predict healthcare outcomes, saving lives and reducing costs by ensuring patients receive timely and sometimes life-saving treatments.

Early last month, the government’s life sciences tsar, Sir John Bell, said that similar techniques could save the NHS billions.

A team at Oxford’s John Radcliffe Hospital is using AI to identify abnormalities in ECG read-outs that could be missed by the human eye, for instance.

The hope is that this system could be used in hospitals across the country to prevent unnecessary hospitalisations caused by false positives, and prevent heart attacks where at-risk patients are sent home because doctors have failed to spot problems.

Another AI system outperformed a panel of experts when asked to diagnose breast cancer based on stained tissue samples, in a separate study published in the Journal of the American Medical Association last month.


Microsoft transforms diagnosis with AI systems

Wax Selection – Leaders in Pharma, Biotech & MedTech Recruitment

The AI system aims to quicken the process of finding illnesses from blood tests.

Diagnosing patients just got a lot easier thanks to a Microsoft partnership with a biotech company that uses AI to check for multiple illnesses.

Taking the next step with technology in the healthcare sector, Microsoft has partnered with biotechnology company Adaptive to create a method for smarter blood tests using AI. The aim of using the technology is to check for hundreds of diseases at one given time, as a universal diagnostic tool.

The use of AI could drastically reduce the amount of time it takes for patients to be diagnosed, by having all types of cases catered for at one point, as well as putting patients through just one trawl of tests instead of multiple. In turn, the technology benefits doctors by saving them the job of carrying out multiple tests, when AI can do it in one swift move.

“We believe deeply in the potential for this partnership with Adaptive and have made a substantial financial investment in the company. We believe deeply in the potential for this partnership with Adaptive and have made a substantial financial investment in the company. With Microsoft’s help, adaptive biotechnologies will attempt to map the genetics of the human immune system or immunome,” Peter Lee, Corporate Vice President, AI and Research at Microsoft said.

Additionally, doctors hope that by checking for more than one disease at once will help with treatments by allowing doctors to look at correlations between different disease states and finding a better cure to target more than one illness.

The system will work by using AI to decode a patient’s immune system, come up with a diagnosis based on past trends and then treat disease with the best medication according to the system.

Adaptive Biotechnologies said to start with the two companies will focus on identifying diseases that are normally diagnosed at a later stage, such as pancreatic and ovarian cancer, and the AI system will help to identify these conditions much more easily and quickly in the future.

“Some conditions, like cancer or autoimmune disorders, can be difficult to diagnose,” said Chad Robbins, co-founder and chief executive officer of Adaptive Biotechnologies. “But this universal map of the immune system will enable earlier and more accurate diagnosis of disease.”

The partnership between the two companies is set to be discussed in further detail at the JP Morgan Healthcare Conference in San Francisco on Wednesday 10th January.