Patient-Centered Clinical Trials USA 2015

Oct 19, 2015 - Oct 20, 2015, Philadelphia

Boost Clinical Trial Productivity – Collaborate with Patients at the Heart of your Research.

Artificial Intelligence – A Brave New World for Pharma

Dr Andree Bates explains how artificial intelligence can help reduce trial time, crunch data, and facilitate healing.

While Stephen Hawking believes that the development of full Artificial Intelligence (AI) could spell the end of the human race, he has also conceded that advances in AI have made life easier for many, including himself.

The scientists and engineers who are producing intelligent machines have made enormous progress in the speed and memory capabilities of their devices, to the point where they vastly exceed those of humans in many respects. However, the intellectual mechanisms of even the most modern robots are still limited to what a designer can program into them in the first place.

Now, with the advent of ‘big data,’ companies are harnessing the power of AI to deliver more focused solutions in a variety of areas; AI helps them understand data in real-time. In a recent interview, Dr Andree Bates, President of Eularis, agreed to share some thoughts on how AI can be applied to the pharmaceutical industry.

Think of how long it took from the beginning to launch Herceptin - the research prior to clinical trials was 10 years, followed by another 8 years for clinical trials. What AI can do is potentially reduce that process down to weeks.

AI crunches research time

“Currently the main uses of AI are in research areas,” says Bates. “AI is far superior to human analyses, as it can analyze vast quantities of data that don’t fit into conventional computers.” It is also believed that the processing powers of AI are superior to other tools. In research on gene mutations, for instance, AI can churn through huge amounts of data and find valuable information.

“Think of how long it took from the beginning to launch Herceptin - the research prior to clinical trials was 10 years, followed by another 8 years for clinical trials,” says Bates. “What AI can do is potentially reduce that process down to weeks.” The implications for the industry could be very exciting, especially with the current focus within pharma being on cutting time and costs spent on clinical trials. In this respect, AI is an area that should be looked at closely in order to enhance return on investment (ROI). 

Can AI surpass doctors?

Pictured in the neonatal intensive <br> care unit at Lucile Packard Children’s Hospital are<br> Anand Rajani, MD; <br>Suchi Saria; Anna Penn, MD, PhD; <br>and Daphne Koller, PhD.By analyzing what is going on within the human body, AI, together with big data, can predict problems even before doctors are able to detect any physical signs. This has resulted in various systems, one of which is PhysiScore, developed by a team led by Dr Daphne Koller and Dr Anna Penn of Stanford University.

PhysiScore uses various data elements to predict whether premature babies are likely to have health issues. A combination of inputs, such as birth weight and gestational age, as well as the real-time data on heart rate, respiration rate, and levels of oxygen saturation, are correlated. This is a task that human beings alone would never be able to manage.

Intersection between AI and Machine Learning

The huge potential of AI is enhanced further by Machine Learning (ML). Bates explains the differences between ML and AI: “AI is the field in which high level mathematical approaches intersect with high level computer approaches, and it really is far beyond anything else for analysis in terms of speed and accuracy. It is essentially computational mathematics.”

Within AI, there are many techniques. AI is the parent technique, and then under that are the more specific approaches. “Machine learning is one field contained within AI,” Bates explains. “Machine learning allows a machine to learn from data without needing rule-based programming.”  

Bates feels that pharma executives should take cognizance of the fact that ML, together with AI, will be critical in driving the future success of their industry. It is, she believes, “The future to drive accelerated growth.” She adds, “How you analyze big data is also critical. By applying the right AI techniques to the right data, it becomes easier to strategize effectively on the overall direction the company wishes to take, identify effective propositions for various segment types within a brand and, particularly when conducting clinical trials, allocate resources and budgets for optimum results.”

AI analytics help profit margins

Bates, who has presented on marketing strategies at more than 45 conferences globally, shares that AI analytics is used by Eularis to enhance pharma clients’ sales and marketing outcomes. “We apply AI-based techniques to everything we do now, although it has taken almost two years of intense research and development - with leading AI Professors and Programmers - to get to the high level we are at, which built on the previous 10 years of development.”

The Eularis team use a variety of techniques within AI, including machine learning and neural networks, as well as deep learning, amongst others. “And, of course, within each area we also use specific approaches for specific types of data,” says Bates, pointing out that AI is a non-linear approach, whereas most other analytical techniques currently in use tend to be linear. “The advantages of AI-based analytics,” she maintains, “are that it understands the non-linearity in data, has far greater speed and accuracy in analyses, and gives a far superior understanding of real driver impacts and their synergistic effects.” 

Where else can pharma use AI?

There are a huge number of uses for AI within the pharma industry. Indeed, Bates believes that, “essentially in any area that requires an intelligent decision, AI can be used more successfully than human intelligence.”  For this reason, relevant project areas include discovery, R&D, clinical trial data analysis, sales and marketing, segmentation, mergers and acquisitions, all operational areas, and company strategy.

R&D is costly and takes many years, but with AI the whole process can be improved and accelerated dramatically.

Bates is enthusiastic about the potential of AI to make significant advances when it comes to helping patients, and even in preventing future epidemics: “We have also been looking at predicting disease outbreaks years in advance. We know that, mathematically, we can do this now. Imagine if we had done this a few years ago, we could have predicted the Ebola outbreak and the companies producing vaccines could have been ready for it.”

The Boston based biotech firm, Berg, is hoping to crack the so-called cancer code soon, a very exciting application for AI. As Bates notes, “R&D is costly and takes many years, but with AI the whole process can be improved and accelerated dramatically.” Berg is also working on identifying biomarkers in diabetes, obesity, and cardiovascular disease.

Robots and AI pave the way for change

Within the next two decades, the way things are done, across multiple sectors of the economy, is likely to change drastically as a result of AI. “Many of the newest leading companies, like Amazon, Apple, Uber, and Airbnb, have embraced it to differentiate,” says Bates. “And whilst it is well known that robots will replace 50% of the manual work in the coming decades, it is also likely that high value activities requiring advanced intellect will be replaceable by AI.” 

Bates explains that there are some surprising new uses being developed for AI in this area. “AI is already being used in law to more successfully identify legal defenses. It is also being used with clinical data to diagnose and predict the progression of cancer cells with more accuracy than human pathologists. I am hard pressed to think of areas where it will not be possible for AI to be a key driver.”

This leads us onto Erica, a famous new carer robot that was developed in Japan. Erica is incredibly human and she shows that even carers roles will probably be taken over by robots in the near future. She is an exciting prototype that demonstrates significant potential in the area of AI. This humanoid robot understands language and uses facial expressions, as well as audio, in conversations. She is the brainchild of Hiroshi Ishiguro, a Professor at Osaka University, who has developed her in collaboration with the Japan Science and Technology Agency, the Advanced Telecommunications Research Institute International (ATR) and Kyoto University. Erica has been created with a blend of Asian and European facial features and speaks Japanese.

In addition to Erica, there is the MEDirobot, whose mission is to provide pain relief for children at various hospitals in Canada and the US. The robot has been found to reduce pain by 50%. Children love interacting with him, according to Dr Tanya Beran, the developer: “We have seen children gain a sense of companionship and even friendship with a robot. They would tell it secrets, play games, and engage in all sorts of activities with it. Moreover, we’ve seen that repeat visits with MEDi increases children’s physical closeness with the machine, wanting to hold its hand and give it hugs.”  

MEDi, an example of AI in action, can have a dramatic impact on individual patients. Beran describes some remarkable cases: “We’ve seen a boy who lost his foot want to get out of bed to practice using his walker just so he could walk down the hallway with MEDi. We have published research demonstrating that children believe robots have the capacity to think and feel in ways that they themselves do. It’s as if they project their experiences and imagination onto them. It has brought tears to many of us to see children’s faces light up when they visit with MEDi despite the overwhelming medical challenges they may encounter.”

It would seem that Stephen Hawking was correct - partly.  With robots to take care of us, AI to crunch the data and find new treatments, and identification of possible health hazards ahead of time, it may well be the end of the world… as we know it anyway.

To learn more about artificial intelligence approaches in pharma sales and marketing, look out for Bates’ upcoming column, ‘Why Pharma Marketers should Consider Artificial Intelligence to Boost the Outcomes from their Marketing Initiatives.’

Patient-Centered Clinical Trials USA 2015

Oct 19, 2015 - Oct 20, 2015, Philadelphia

Boost Clinical Trial Productivity – Collaborate with Patients at the Heart of your Research.