Exponential technologies are proving to be disruptors in medicine and healthcare delivery. Dr David Jankelow describes how innovation is an enabler of more efficient and quality care.
The South African Heart Association 2017 Congress was held in November last year at the Sandton Convention Centre. It was a highly successful meeting, with almost a thousand delegates, who were educated and stimulated by an eminent local and international faculty, "the rock-stars of cardiology". The congress theme was "Fundamentals to Innovation" and showcased developments in every aspect of cardiovascular medicine.
Disruption throughout history
Who was he? An Italian explorer and navigator, who sailed across the Atlantic in 1492 in the hope of finding a new route to the East. He failed in this endeavour, but in the process became one of the most important spice traders of the fifteenth century. These men - the traders - yearned, burned and travelled the world for cinnamon, nutmeg, cloves and pepper. At the time, it was widely believed that seasoning preserved meat. It didn't; it merely disguised the taste when the meat had gone rotten!
The spice trade was the world's biggest business. It established and destroyed empires, it led to the discovery of new continents and in many ways laid the foundation for the modern world.
But then, the price plummeted and the industry collapsed from its former glory. It was described as a slow-moving train wreck that nobody saw coming. The reason? The spice trade was disrupted by the ice trade!
Frederick Tudor and subsequent refrigeration
Today we take ice-cold drinks for granted, but two hundred years ago it was a luxury only the wealthy could afford. Frederick Tudor became known as the "Ice King". As a young man, he had looked out at the frozen Boston River and wondered whether the floating ice could be used to preserve fresh foods. He invented a horse-drawn ice cutter, which scored the ice into giant blocks that could be packed tightly to minimise melting. Ice cut in New England was exported across the globe on insulated ships (with sawdust and hay) to tropical lands, where it was used to preserve meat, fruit and vegetables. It was also used to soothe the suffering of those with yellow fever.
Interestingly, none of the previously prosperous spice companies crossed over to the ice trade. They didn't even offer the use of their ships to transport the ice. Nobody made the change.
Frederick Tudor's made a cold fortune, but eventually his booming enterprise was disrupted as well. The industrial ice-maker was created. A block of ice would be delivered with your morning milk to be placed in a specialised cupboard, in which fresh foods were stored.
By the end of the 20th century, little remained of the Tudor Ice Company, which had desperately tried to convince the world that natural ice was of a much better quality than that manufactured industrially. Nobody, however believed this traction.
The subsequent development of the cheap electric motor led to the creation of the fridge, the Kelvinator, and one now could make ice at home. So the ice industry was disrupted by the ice cupboard, which then was disrupted by the fridge. So it should be obvious - Could anything disrupt the fridge?
So what is the relevance?
We can learn lessons from these stories about the evasiveness of disruption and how difficult it is to foresee. In each case, large businesses involved in one version failed to make the transition to a new model. The disruption took place over years - they had time to adapt, but they did not.
What are the implications for medicine and healthcare?
Are you ready for a technological revolution in medicine? It will be exponential, it won't be stopped and we need to embrace it.
The tools of our trade are the decades old stethoscope and sphygmomanometer. Are they not almost obsolete?
We simply don't like change and it is no different in medicine, which has always had a complicated relationship with technology - embracing some innovative tools but resisting others. But like it or not, digital disruption will happen.
Perhaps the next generation of doctors will have to be trained in an extra, different set of skills that will equip them for both clinical and digital challenges.
What will the future look like?
With the exponential rise of technologies such as artificial intelligence, genomics, nanotechnology, telemedicine, virtual reality, robotics, 3D printing and even blockchain, healthcare is going to change in its entirety!
The rise of the smartphone, apps, trackers and health sensors
We can help people live healthier lives by reforming access to our own personal health data. It is also likely that the smartphone will become the medical hub of the future, allowing us to monitor our own health. It is projected that there will soon be more than six billion subscribers globally, half of whom will make use of some form of health app.
We are also now in an era of portable diagnostics enabled by sensors in wearable devices.
Digital apps are vying to diagnose everything from skin cancer and concussion to Parkinson's disease. For complex diseases such as diabetes, apps can change the way patients cope by monitoring blood glucose level and even food intake.
Although these technologies leverage insight into one's own self, without our, the medical professions supervision, patients could fall prey to misinformation.
Shifting the focus from sick care to health care
Patients are becoming engaged and empowered. They increasingly use the internet to research their symptoms and make their own medical decisions. Trust in doctors is also down to an all-time low. We need to shift the focus from sick care to health care, which will not advance, unless physicians allow patients, in part, to help themselves.
Self-ownership of health data is on the rise
A fundamental problem is that patients lack knowledge and control. Access to medical information not only offers both, but also improves the efficiency of care.
Although health records are increasingly electronic, they are often still trapped in silos, and many medical errors can be traced to poorly coordinated care. With data at our fingertips, common standards to enable sharing and a strong incentive to get things right, errors can be identified earlier.
Many countries are now opening up access to medical records, but few have gone as far as Sweden. This country aims to give all its citizens full electronic access by 2020. Studies show that such patients, have a better understanding of their illnesses and their treatment being successful.
A recent Economist article (How data will transform health care) concluded that, "No one has a greater interest in your health than you do. Trust in Doctor You".
We now have the processing power to analyse massive amounts of data at lightning speed. Supercomputers such as Google DeepMind and IBM Watson are the crème de la crème of artificial intelligence (AI).
DeepMind was co-founded in 2010 by Demis Hassabis, a child genius and chess prodigy who went on to study computer science at Cambridge University and neuroscience at University College London. DeepMind's mission is to solve intelligence, sort out the most vexing global problems, and then make the world a better place.
What is AI?
It is the science of making machines smart. These systems use programmatic computing, natural language processing and, most importantly, deep machine learning so the system develops enormous expertise. In essence, the computer is trained to figure out a problem rather than having the questions and answers programmed into it. AI has the potential to deliver the power to unlock the global world of unstructured data, ninety per cent of which was created in the past two years and will amount to many zettabytes of information. A zettabyte is the equivalent of a trillion gigabytes!
Eighty per cent of the world's data is unstructured. The most common form is text, generated and collected in a wide range of forms, including MS Word documents, email messages, PowerPoint presentations, survey responses, transcripts of call centre interactions, as well as blogs and social media posts. Other forms include images, audio and video files.
How much do we contribute to this?
What happens on the internet every minute - over two hundred million emails sent; more than a million Google searches and YouTube views; five hundred thousand Facebook logins; a hundred thousand tweets; twenty thousand hours of video and sixty thousand hours of music downloads. A lot of noise is created online in sixty seconds. The challenge is to see and understand it and, perhaps more controversially, to control it.
Organisations and institutions use structured data through databases and spreadsheets so that patterns can be easily identified. However, unstructured information represents a significant source of opportunity. We need to be able to understand what's in all of this or risk missing out on significant digital intelligence. It is however difficult to gain insight with conventional methods. Cognitive data analytics will make it possible to make the right decisions faster, with sound statistics and facts. Harnessing unstructured "noise" will help businesses to stay ahead of the competition.
When humanoid robot NAO, powered by IBM's Watson, opened a symposium at the Oxford Union, he introduced the speakers and explained, "Watson has enabled me to learn in ways that I could not have done before. Not only can I read up to eight hundred million pages a second, I also understand natural language better than ever before."
Cognitive systems will be able to see, hear and understand just like humans do. In this new age of augmentation, your capabilities will be far advanced by systems that help you think, robots that help you make, and a digital central nervous system that connects you to the world, far beyond your natural senses. It's pretty amazing stuff!
A watershed moment in AI occurred in 1997. This was the first defeat of a reigning world chess champion by a computer under tournament conditions. IBM's Deep Blue, a ten million dollar machine, beat Garry Kasparov, in what Time magazine called "the match of the century". Deep Blue was able to play chess by brute force. But was it intelligent? The answer, categorically, is "No". Other than play chess, Deep Blue could not perform any other tasks. It would even have failed at naughts and crosses. Garry Kasparov, however, can speak a number of different languages and do many other things. Deep Blue was therefore extremely narrow AI, the realms of which have since evolved immensely.
Now, for the first time in history, we're seeing the development and demonstration of general-purpose AI.
In 2011, a revelation was witnessed on the television quiz show Jeopardy. IBM's Watson demolished the most successful players in the history of the game. It didn't smile or show any emotion; it just kept giving the correct answers. The audience was left in shock.
The ancient Chinese game Go is the world's oldest continuously played game. It is one of the simplest but also the most abstract, as the possible configurations exceed the number of atoms in the universe!
Recently AlphaGo, an algorithm developed by Google DeepMind, showed that a machine could approximate human intuition, outsmarting the best human brain in this game. It beat the world champion Lee Sedol, in a million-dollar challenge. Sedol is to Go what Roger Federer is to tennis. AlphaGo rapidly developed grandmaster skills, playing in an unconventional way. Scientists hadn't expected such a feat for at least another decade. It is a giant leap for computer science, showing that AI can truly learn and think in a human way.
AlphaGo's creators say the algorithm can learn many more things without alteration or guidance, and it will be truly general purpose.
AI has the potential to influence everything we do - from the way we answer our phones to the way we drive our cars or perhaps even the way our cars will drive us.
These events were not planned merely for entertainment value or to just win board games. AI has rather shown that it just might impact on every aspect of humankind, including medicine and health care.
It changes everything - including medicine
Oncology: At Memorial Sloan Kettering, oncologists are now using Watson in the form of an iPad app. Massive amounts of published oncology studies have been loaded into the cloud-based system. A patient's clinical, biopsy and imaging information is entered and within seconds, the app offers a number of treatment plans, with supporting evidence and likely success rates for each treatment option.
The system will examine thousands of textbooks, guidelines, and several hundred thousand other journal articles, clinical trials and documents. It offers the ability to extract precise answers from a staggering amount of information. Quite incredible, you will agree! Watson for oncology has been in development for six years. It now offers guidance for the treatment of the twelve cancers that account for eighty per cent of the world's cases.
Radiology: A role is also emerging for machines as intelligent assistants to radiologists, who often have to analyse thousands of images a day. With this kind of volume, radiologists may suffer eye fatigue and decision making then becomes harder. Cognitive radiology uses clinical knowledge and advanced analytics. It is able to detect anomalies that aren't perceptible to the human eye. AI won't replace radiologists; however, radiologists who use AI, will replace those who don't. This will be relevant to all professions.
Skin cancer detection: Every year, there are more than five million new cases of skin cancer in the USA. While the five-year survival rate for melanoma detected in its earliest stages is around ninety-seven per cent, the cure rate drops significantly if diagnosed at a later stage. Earlier detection would therefore have an enormous impact on outcomes.
Universal access to health care was on the minds of computer scientists at Stanford when they developed an AI algorithm for diagnosing skin cancer. They used a database of one hundred and thirty thousand skin disease images and trained their system to diagnose potential cancer. From the very first test, it performed with inspiring accuracy.
It was fed each image as raw pixels with an associated disease label. It's performance was measured through the creation of a sensitivity-specificity curve, where sensitivity represented its ability to correctly identify malignant lesions and specificity represented its ability to detect benign lesions. The algorithm matched the performance of all tested experts, demonstrating an intelligence capable of classifying skin cancer with a level of competence comparable to twenty-one Stanford dermatologists (reported in Nature).
The plan is to develop a smartphone app, bringing reliable skin cancer diagnoses to our fingertips. Outfitted with deep neural networks, mobile devices could potentially extend the reach of dermatologists and provide low-cost, vital diagnostic care; in essence "a pocket dermatologist".
A major issue is information overload, the volume of which is growing at such a rate that it is impossible to keep up to date. A new book is published every thirteen seconds; a research paper every thirteen milliseconds; and an interesting posting on the internet, every thirteen microseconds. To add to the complexity, biomedical data is doubling every five years. Our only saviour will be AI. Imagine an individual who can read an unlimited number of documents, and understand and completely retain every bit of information. Imagine you could ask that person an intuitive question; that's essentially what AI gives you. It represents a way to look at vast collections of data at scale and then extract the needle in a haystack. No single doctor could match this kind of ability, which will discover relationships that may not have been obvious.
But it goes even further, to the analysis of large data sets that can't be analysed, searched, interpreted or stored using conventional processing methods. This includes mobile phone applications, wearable technology, social media, environmental and lifestyle-related factors, sociodemographics, "omics" and electronic health records. Whereas big data by itself is useless, using AI to make predictions or decisions, it has the potential to transform current clinical practice into personalised precision medicine.The future of work
You might be wondering whether we could all eventually be replaced by intelligent automation. This was the main topic of a recent issue of the Financial Mail, "Robots are coming for your job: What the future of work in SA looks like".
According to The World Economic Forum, no less than forty-one percent of work activities will be at risk for automation. This doesn't mean that these jobs will disappear - it means that individuals will need to adapt to new ways of doing things. About eight to nine percent of jobs will disappear and be replaced by AI and robots, which is a social challenge to face up to. Workers will therefore have to be tech-savvy. Core skills required for all occupations will be different to what is needed right now.
Man and machine
Garry Kasparov has said that few people in the world know better what it's like to have your life's work threatened by technology. He ended a fascinating TED talk, Don't fear intelligent machines; Work with them, with the following:
"Machines have calculations. We have understanding. Machines have instructions. We have purpose. Machines have objectivity. We have passion. We will need the help of intelligent machines to turn our grandest dreams into reality. And if we fail, it's not because our machines are too intelligent, or not intelligent enough. If we fail, it's because we grew complacent and limited our ambitions. Our humanity is not defined by any skill. There's one thing that humans can do. That's dream. So let us dream big!" I urge you all to dream big in medicine.
Augmented intelligence in medicine - our valued assistant
Divya Chander, physician and neuroscientist of Stanford University has said "I would trust a human plus AI more than a person alone." The question however is, at what stage will you trust the AI by itself?
Augmented intelligence will not only be an enabler, it will also enhance human expertise. It will free us up from repetitive tasks so that we can get back to the actual business of caring for and treating our patients. Isn’t that what it is all about? Machines will never be able to replace human caring and empathy. Rather, they will take the robot out of the doctor!
A crisis in South Africa: cardiac specialist training and a shortage of health professionals
Professor Karen Sliwa, Director of the Hatter Institute for Cardiovascular Research in Africa, has published on this: "There are approximately two hundred cardiologists for fifty two million South Africans (1 per 260 000). If you compare the situation with Brazil or other BRIC countries with a similar disease burden - Brazil has eight thousand cardiologists for a population of one hundred and eighty five million (1 per 23 000) or just over ten times more than in South Africa."
How do we deal with this crisis in the medium term? Telemedicine might be one option to improve specialist care in remote areas and hospitals.
Meet the InTouch telemedicine robot, to which devices such as stethoscopes or ultrasound probes can be attached. It can be controlled remotely by a specialist anywhere in the world to provide urgent clinical advice. Robotic technology like this could also address the shortage of intensivists that exists all over the world. It will offer another set of eyes to watch over our critically ill patients.
CALL TO ACTION; ADVOCACY
My appeal to all of you, the physicians, cardiologists, cardiac surgeons, nursing profession, allied cardiac professionals, funders, policymakers and regulators. I urge you to help ensure that all the future innovations in medicine be made available to all the people of beautiful South Africa. Health care must be equally available to all!
- https://www.economist.com/.../21736138-welcome-doctor-you- revolution- health-care-coming
- Esteva, A. et al. Nature. 2017; 542; 115.
- Krittanawong, C. et al. J Am Coll Cardiol. 2017; 69(21); 2657.
- Ziady, H. Financial Mail. 2017; October 12; 20.
- https://www.ted.com/talks/garry_kasparov_don_t_fear_intellig ent_machines_work_with_them?utm_campaign=tedspread&u tm_medium=referral&utm_source=tedcomshare
- Sliwa, K. et al. Cardiovasc J Afr. 2016; 27: 157.
- David Jankelow, MBBCH, FCP (SA), FACC Congress Chairman, SA Heart 2017.