Watson is better in Jeopardy than humans
In 2011, people witnessed an interesting competition on the television quiz show Jeopardy. It featured the two best players in the history of the show, Ken Jennings, who had the longest unbeaten run of 74 winning appearances, and Brad Rutter, earner of the biggest prize of $3.25 million. Their opponent was a huge computer with over 750 servers and a cooling system stored at a location so as not to disturb the players. The room–sized machine was made by IBM and named after the company’s founder, Thomas J. Watson. It did not smile or show emotion, but it kept on giving good answers. At the end, Watson won the game with $77,147 leaving Rutter and Jennings with $21,600 and $24,000 respectively. It left the audience in shock and awe at the same time.
Cognitive computers have been developing rapidly over the last few years following three technological breakthroughs. One is cheap parallel computation due to a new kind of chip called a graphics processing unit (GPU). The second one is accessible big data due to massive databases, web cookies, wearable devices and decades of search results. The third one is building better algorithms due to the services of Netflix, Google, Amazon and the others.
Do we have to fear cognitive computers taking our jobs?
Some started to contemplate about how A.I. would replace many jobs, overtake the human race in thinking. Stephen Hawking even said that the development of full artificial intelligence could spell the end of the human race. Elon Musk agreed.
However, it is not inevitable that the use of A.I. leads to the loss of the human touch. In 1997, IBM’s supercomputer Deep Blue could beat Garry Kasparov, the reigning chess grand master that time. He said he could have performed better if he had access to the same databases as Deep Blue. So later, freestyle matches were organized in which supercomputers could play against human chess players assisted by AI (they were called human/AI centaurs). Guess what! In 2014 in a Freestyle Battle, the AI chess players won 42 games, but centaurs won 53 games. The best potential pair is a human with technology. This is the only balance that can lead to a positive future with more and more disruptive innovations including ever-improving cognitive computing but an also ever-improving human intelligence and wisdom. This is the winning combination.
From Stethoscope to Cognitive Computers
People don’t like change. It is not any different in healthcare. It took plenty of time until the stethoscope, the symbol of healthcare was accepted by the medical community back in the 19th century. The instrument was invented by French physician René-Théophile-Hyacinthe Laënnec, who published its description in 1819, but it took several decades until doctors actually started using it.
I don’t expect any difference with A.I. After many of my talks, physician colleagues ask me whether artificial intelligence might replace them in their jobs and whether algorithms can eventually become better at making diagnoses. Both will happen but not the way they imagine it. Huge waves are coming to healthcare to transform the job of physicians into something distinctly different than before. Although some of their tasks will be taken over by A.I., they will have more time for others, for example, deal with patients with real care and patience. Doctors will not have to struggle with being up-to-date in medical research, with administrative tasks, with consultation or making notes. They do not have to have a headache about how to choose the best therapy.
Cognitive computers will help physicians diagnose much better – the same way stethoscope could change the medical profession from the early 19th century on; when a wooden tube working like an ear trumpet could make doctors listen to cardiac and lung sounds at the point of care.
A.I. will open a new dimension for healthcare
What even the most acclaimed professors know cannot match cognitive computers. As the amount of information they accumulate grows exponentially, the assistance of computing solutions in medical decisions is imminent. A.I. will open new dimensions for doctors on the personal level as well as for hospitals and other medical institutions on the structural level.
On the institutional level, the most obvious use of A.I. will be data management. Collecting it, storing it, normalizing it, tracing its lineage – it is the first step in revolutionizing the existing healthcare systems. Recently, the AI research branch of the search giant, Google, launched its Google Deepmind Health project, which is used to mine the data of medical records in order to provide better and faster health services.
It could also analyze entire healthcare systems. For example, 97 percent of healthcare invoices in the Netherlands are digital containing data regarding the treatment, the doctor, and the hospital. These invoices could be easily retrieved. A local company, Zorgprisma Publiek analyzes the invoices and uses IBM Watson in the cloud to mine the data. They can tell if a doctor, clinic or hospital makes mistakes repetitively in treating a certain type of condition in order to help them improve and avoid unnecessary hospitalizations of patients.
A.I. will help doctors in previously unimaginable ways
Precision medicine, targeted treatments, and personalized solutions: these are the buzzwords in current healthcare – and not by chance. A.I. will help physicians work out the best therapies for their patients. IBM has already taken the first steps. Watson launched its special program for oncologists – and I interviewed one of the professors working with it – which is able to provide clinicians evidence-based treatment options.
Moreover, A.I. has the potential to eliminate hideous tasks, such as administrative or repetitive work. For example, IBM’s Medical Sieve will become the next generation “cognitive assistant” with analytical, reasoning capabilities and a wide range of clinical knowledge. The algorithm will be qualified to assist in clinical decision making in radiology and cardiology. A.I. assistants could also look up relevant medical information and keep doctors up-to-date in clinical research. IBM has the capacity to scan through millions of pages in seconds. Imagine the possibilities that lie ahead of us!
Moreover, A.I. could eradicate waiting time by optimizing both physicians’ and patients’ schedules. It could prioritize doctors’ emails so as the most urgent messages would reach them in time. It could also help patients sort out their simpler medical issues thus reducing the burden on doctors. If you look at all the advantages of A.I., your question will only be when it will finally reach our hospitals and why it is not there already.
A.I. will augment clinical practice
There are already certain places where A.I. is tested. Google DeepMind already launched a partnership with the UK’s National Health Service to improve the process of delivering care with digital solutions. In June 2017, DeepMind expanded its services – first of all, its data management app, Streams, to another UK hospital. This expansion comes despite ongoing controversy over the company’s first NHS data-sharing agreement.
Google DeepMind’s number one competitor, IBM Watson is used at the Alder Hey Children’s Hospital as part of a science and technology facilities council project being run by the Hartree Centre. Dr. Iain Hennessey, Clinical Director of Innovation told me, the UK is spending £300 million pounds to develop its capabilities in this area and they are one of the first use cases.
I also talked to Martijn G.H. Van Oijen, Ph.D. is an Associate Professor at the Academic Medical Center at the University of Amsterdam about Watson. As a clinical epidemiologist, he worked with IBM Watson for Oncology in preparation for a research proposal studying the role of Digital Decision Support tools. With several clinical colleagues of the department of Medical Oncology, they studied Watson for Oncology’s approach in approximately 400 surrogate patients with breast, lung or colorectal cancer. He believes that Watson for Oncology could result in a reduction of costs and efforts.
They both agreed that AI cannot be a substitute for communication. Right now, the technology is in its infancy, but improving all the time, said Hennessey. My prediction is that it will gradually appear in more and more hospitals and get through the initial phase very quickly.
What can we do to facilitate change?
Although A.I. is coming as fast as Daenerys Targaryen to Westeros, it will need the medical community’s understanding, initiative and drive for a better healthcare in order to work its best. If we introduced A.I. everywhere from Bolivia to Harvard, healthcare would not change a bit. There are many reasons for it and every stakeholder in healthcare should play its part to improve those elements.
- Medical professionals should acquire basic knowledge about how AI works in a medical setting in order to understand how such solutions might help them in their everyday job. They also need to constantly think about where automation could let their jobs do better.
- Decision makers at healthcare institutions should do everything to be able to measure the success and the effectiveness of the system. This is the only way to assess the quality of AI’s help in medical decision making.
- Companies such as IBM should communicate even more towards the general public about the potential advantages and risks of using AI in medicine – and sort out data privacy issues such as the recent concerns with the NHS!
- Non-English speaking countries should invest in natural language processing (NLP). If the patient information is not in English, A.I. needs to understand the content and context of the structured and unstructured information in that language.
Artificial intelligence will become the stethoscope of the 21st century, but it can only do its best when there is a wide cooperation between technology and the medical community. Let’s make it happen together!