Tuesday, November 8, 2016

Update on Artificial Intelligence and Healthcare

The sheer volume of available medical knowledge has long since outstripped even the most capable clinician's ability to review and properly analyze the all the data now being generated and collected about a patient's health. Today it requires the assistance of sophisticated computer systems to not only help them with the analysis of the data, but to also stay on top of breakthroughs in genomics, predictive analytics, clinical decision support, population health management, and the continuous changes in best practices. Read How Healthcare Can Prep for Artificial Intelligence.

Even the best, most attentive health care providers simply can’t stay on top of all the data and the growing body of ever-changing medical knowledge. Doctors just can't be there for a patient around the clock, to catch every bad habit or reinforce every positive behavior. However, Artificial intelligence (AI) has evolved to the point that it can now begin to significantly augment the healthcare provided by physicians, pharmacists, nurses, and health coaches – not to mention self care.

Today, over 90 Artificial Intelligence (AI) startup comanies are already active in developing new solutions in the areas of Imaging & Diagnostics, Drug Discovery, Remote Patient Monitoring, Hospital Management, Customer Service, Healthcare Analytics & Research, Patient Alerts & Reminders, Telemedicine, Public Health, and more. See Artificial Intelligence Startups in Healthcare by CB Insights.

Selected Quotes

  • In a recent interview, AthenaHealth CEO Jonathan Bush noted the limitations of traditional doctors and said, “The human is wrong so freaking often, it’s a massacre.”
  • “By 2025, AI systems could be involved in everything from population health management, to digital avatars capable of answering specific patient queries.” — Frost & Sullivan.
  • Stephen Hawking has said the development of full Artificial Intelligence (AI) could spell the end of the human race – and Elon Musk agreed.

Current Examples of AI in Healthcare

Aside from the big players like IBM and Google, some of the other major startup companies focused on AI in healthcare include: Ayasdi, Babylon Health, Digital Reasoning, Gauss Surgical, H2O AI, iCarbonX, Lumiata, Pathway Genomics, Stratified Medical, Welltok and Zephyr Health – to name just a few.

The following are specific examples of major AI systems and project activities in healthcare that you might want to explore further:
  • Google Deepmind Health project is being used to mine the data of electronic medical records (EMR) in order to provide better and faster health services. 
  • Medical Sieve, an ambitious long-term exploratory project by IBM is being used to build a next generation “cognitive assistant” with analytical, reasoning capabilities and a wide range of clinical knowledge to assist in clinical decision making in radiology and cardiology.
  • IBM has also launched Watson for Oncology that is able to provide clinicians with better evidence-based treatment options. 
  • Babylon Health has launched an app this year which offers medical AI consultation based on personal medical history and common medical knowledge.
  • Medical start-up SenseLy uses AI and machine learning to support patients with chronic conditions in-between doctor’s visits using Molly, the world's first virtual nurse. 
  • Deep Genomics aims at identifying patterns in huge data sets of genetic information and medical records, looking for mutations and linkages to disease. 
  • Human Longevity offers patients complete genome sequencing coupled with full body scan and very detailed medical check-up to help spot cancer or vascular diseases in their very early stage. 
  • Baidu has introduced an artificial intelligence-powered chatbot called Melody to connect with patients, field medical questions, and suggest diagnoses to doctors. It is a new feature of the Baidu Doctor app it launched last year.
  • Mount Sinai Health System has tapped CloudMedx to help pinpoint people at risk of congestive heart failure as part of its emerging program dubbed HealthPromise.
  • The Computerized Patient Record System (CPRS) developed and implemented by the Veterans Health Administration (VHA) contains a number of modules that use first generation AI capabilities, e.g. Clinical Alerts, Clinical Reminders. 
  • Open AI is a non-profit AI research company that aims to collaboratively develop and promote carefully an open source AI platform and software to benefit humanity as a whole. 
  • VA Collaborating with Flow Health to Bring AI and Precision Medicine to Veterans. Flow Health has formed a five-year partnership with the US Department of Veterans Affairs (VA) to build a medical knowledge graph with deep learning to inform medical decision-making and train artificial intelligence (AI) to personalize care plans.

AI Market & Financial Investment
According to a recent 2016 report by Frost & Sullivan, use of AI is growing in healthcare, with the market poised to reach $6 billion by 2021, up from $600 million in 2014. Cognitive solutions such as IBM’s Watson system are capable of sifting through huge volumes of data and providing guidance and decision support to healthcare providers, thereby improving work flow and patient care.

IBM is serious about bringing Watson into the healthcare industry. To build Watson’s medical credibility, IBM has spent $4 billion dollars purchasing companies that had huge stores of medical data, from billing records to MRI images. Read IBM's Watson Recommends Same Treatment as Doctors in 99% of Cancer Cases.

Artificial Intelligence (AI) coupled with the Internet of Things (IoT) could be the answer to solve major healthcare challenges that the world is facing today. IoT can helping care to move from hospital to home in low acuity and post-operative scenarios, with wearable sensors being monitored and analyzed continuously in real time by AI systems. The number of connected IoT healthcare devices is estimated to be $2.5 trillion by 2025. For more detail, read IoT and AI: Potent combo redefining healthcare.

Key Issues

Artificial Intelligence (AI) technology already exists, and is being deployed in many other arenas, such as finance and business intelligence. But in health care, there’s a human and psychological barrier: can we trust AI with our health?

Some of the key issues that need to be addressed over the coming decade before Artificial Intelligence (AI) systems are widely deployed include:
  • Ethics - This emerging issue is concerned with the moral behavior being programmed by humans into robots and other 'smart' AI systems, e.g. Roboethics.
  • Privacy & SecurityWhen AI systems are turned loose to monitor health information systems gathering data on all facets of your personal health, concerns about who has access to the data and who it is being shared with are just a few of the issues that must be adequately addressed upfront.
  • Jobs - The Bank of England has predicted that intelligent machines might take over 80 million American and 15 million British jobs, respectively over the next 10 to 20 years. The healthcare industry will not be immune to this change.
  • Legal Issues - One of the most important points of interest that needs to be hammered out first is the legality of these machines. When a doctor's gross negligence leads to a misdiagnoses and patient harm, the fault is placed squarely on the shoulders of the offending physician. What happens if an AI system were to misdiagnose a patient? Who's to blame?
  • Open Source - Many new 'open source' tools are arriving that can now run on affordable hardware and allow individuals and small organizations to perform prodigious data crunching and predictive tasks. Read about H2O, OpenAI, and other machine learning and AI tools being used in healthcare at Open Health News.

Conclusions & Next Steps

When and how artificial intelligence (AI) systems will stand in for doctors, nurses, therapists, or specialists is yet to be determined. The use of AI systems is already fairly prolific in some other industries. But the consensus is that these systems aren’t yet sophisticated enough for use in healthcare. However, expect all major electronic health record (EHR) systems of the future to include an AI platform within their architectural structure.

At this stage, AI is perceived as one of a group of emerging technologies that will increasingly be used to augment human capabilities. However, within 10 years AI coupled with IoT, Blockchain technology, and 'Big Data' Analytics may actually be used to carry out many of the functions currently being performed by clinicians. Read Making Us Better At The Things We Do Best.

As we continue to move forward with further development and deployment of AI systems, the following are some of the next steps that need to be taken:

  • Ethical rules and standards to be incorporated in AI systems used in the healthcare need to be extensively addressed.
  • The health IT industry needs to take careful, incremental steps when developing and implementing AI systems in healthcare over the next decade to avoid causing any harm.
  • Medical professionals need to become much more knowledgable about AI technology and systems and work closely with developers before we move forward much further.
  • Patients need to get accustomed to interacting with AI systems, discovering the potential benefits to themselves – to trust these new systems.
  • Companies developing AI healthcare solutions need to carefully communicate with the general public about the potential advantages and risks associated with the use of AI in medicine.
  • Healthcare oversight institutions need to carefully measure the success, shortcomings, and effectiveness of AI systems before they are widely deployed over the coming decades.

The next few decades are going to be an exciting time in healthcare. You might want to take a few minutes to read Searching for Godot – Health 4.0 by 2040.



Selected Links


Friday, October 14, 2016

Searching for Godot – Health 4.0 by 2040

1.0 – Introduction to Health 4.0

Health 4.0 focuses on collaboration, coherence, and convergence – or connecting all available health information, services, devices and people together in a more meaningful way. See Global Artificial Intelligence Network for 2040.

Future Scenario: By 2040, a space-based global artificial intelligence (AI) network of satellites will be put in place that will monitor and help provide healthcare to people on Earth and in colonies across our solar system on the Moon, Mars, and other locations. The system will be linked to massive global health data warehouses storing data from a wide range of health IT systems, e.g. Electronic Health Record (EHR) systems, Personal Health Records (PHR), Health Information Exchange (HIE) networks, wearable fitness trackers, implantable medical devices, clinical imaging systems, genomic databases and biorepositories, surgical robots, health research knowledgebases and more.

The space-based global AI system will monitor and analyze the health data gathered on all humans in real-time, detecting potential individual and public health issues. The global AI system will detect problems, diagnose them, send alerts to patients and their healthcare providers, and recommend treatment plans to resolve the healthcare issue. The system will also be interfaced to pharmacies, laboratories, health insurers, public health agencies, and other institutions as needed. The system will also be able to monitor a patient's progress, as well as adherence to recommended treatment plans. It will also seek to anticipate potential healthcare issues and provide preventive health and predictive health information tailored to each human.

Evolution of Health IT Systems


The following is a brief overview of the evolution and use of health information technology in the US since the late 1960's:

  • Health 1.0 (1970's – 1990) = Modular Health IT Systems, e.g. Patient Registration, Billing, Pharmacy, Lab…
  • Health 2.0 (1990's – 2010) = Integrated Electronic Health Record (EHR) Systems + Personal Health Records (PHR) + Clinical Imaging
  • Health 3.0 (2015 – 2030) = Networked EHR Systems + Genomic Information + Wearable & Implantable Sensor Data
  • Health 4.0 (2030 – 2040's) = Global Networked EHR Systems + Artificial Intelligence + Convergence of all Technologies Above + Real-Time Global Data Collection & Analysis

As we approach 2020, we are currently in the process of developing, and implementing Health 3.0 technologies and solutions in the US and other advanced nations across the globe. Preliminary design and pilot testing of some Health 4.0 solutions is just beginning.

Looking at 2040 and Beyond

Looking ahead to when Health 4.0 systems will finally start rolling into place, keep in mind some of the following predictions for Healthcare and Health IT systems by 2050. See Health 2020-2050.

  • By 2050 we will see more instances of global pandemics and the spread of deadly diseases as a by-product of the skyrocketing growth and migration of the global population.
  • Rise of 'Regenerative Medicine', Genetic Engineering, Stem Cell Research, and the development of 'Human Augmentation' technologies will dramatically alter people's life spans and capabilities.
  • Use of biorepositories and genomic information systems will further transform healthcare and help lower costs.
  • Emergence of future knowledge driven global health platform and solutions will be based on 'open' standards and technologies.

Selected Links




2.0 – Building a Global Artificial Intelligence Platform for Health 4.0

Futurist Ray Kurzweil has predicted that computers will be as smart as humans by 2030. By 2045, he claims 'artificial intelligence' systems may be a billion times more powerful than our unaided 'human intelligence'. Are you prepared for what this means?

In computer science, an ideal Artificial intelligence (AI) system is designed to mimic cognitive functions that humans associate with other human minds, such as “analyzing”, "learning" and "problem solving". Cognitive computers are self-learning artificial intelligence (AI) systems that can find patterns in massive collections of unstructured data and turn it into a presentable form for many others to use. For more detail, see Wikipedia.

The promise of Artificial Intelligence (AI) has always been just beyond the horizon, not quite realistic yet still visible to our imagination through movies and literature. At its inception, AI was initially deployed for highly selective defense or space exploration applications. However, over time it has steadily advanced and has begun to be utilized in many other industries, such as healthcare and manufacturing.

Artificial Intelligence (AI) and Data-Driven Medicine

Artificial Intelligence (AI) and data-driven medicine are the next frontier in the healthcare revolution. Electronic Health Record (EHR) systems have been widely adopted by all major healthcare institutions across the US over the past decade. Now state and local Health Information Exchanges (HIE) networks are being deployed to allow access to data in the EHR systems by healthcare providers whenever or wherever it's needed.

In the meantime, consumers have been gradually starting to use Personal Health Record (PHR) systems that contain their patient data from the various EHR systems of healthcare institutions where they have been treated. Now we are seeing the growing use of other healthcare technologies that are also gathering and generating even more personal health data, e.g. wearable and implantable devices, genomic information systems and biorepositories, clinical imaging systems, and more.

It turns out that Medical data is the essential part of today's comprehensive healthcare systems. However, processing and analyzing the massive quantity of data now being generated by the wide range of converging health information technologies is almost too much to handle. It is an area the healthcare industry is struggling to come to grips with as it turns more and more to use of Artificial Intelligence (AI) as the means to gain control of the growing mountain of health related data.

Roughly every three years, the amount of medical data on the planet doubles in size. By 2020, it is expected to double every 73 days.

It is time to begin focusing more proactively on the design and development of a 'global healthcare platform for 2040' built on artificial intelligence (AI) technologies.

Artificial Intelligence (AI) in Healthcare

Analysts predict a tenfold growth of the use of artificial intelligence in healthcare in the next five years, for everything from cancer diagnosis to diet tips. According to Frost and Sullivan, healthcare providers will spend $6 billion per year on artificial intelligence tools by 2021. Google, IBM and Microsoft are all investing heavily in healthcare and analysts predict 30 percent of providers will run cognitive analytics on patient data by 2018. See Artificial Intelligence: There's Still Hope for the Human Race

Early types of these type of cognitive systems built on artificial intelligence (AI) technologies have already started entering the market. These include 'smart' triage systems that check patients’ symptoms against massive health data warehouses, then advise patients and providers what they should do next. Artificial Intelligence (AI) systems are also being used to help consumers when buying health insurance, to monitor biometric data from personal fitness trackers, analyzing genomic data to predict and potential life-threatening diseases, and much more.
Artificial Intelligence (AI) is advancing rapidly and is in the process of transforming the face of healthcare. Just a few of the many areas in which AI is being used to affect practice management and healthcare services include Diagnosis and Treatment, Disease Management, Personal Health & Wellness, Utilization Management & Reimbursement. Read 5 Ways AI is Changing Healthcare.

Other areas where AI technology and data-driven systems can be designed, developed, and used to improve healthcare include:

  • Examining and analyzing genomic data on hundreds of millions of patients.
  • Building systems that gradually teach themselves to become more accurate in its diagnosis.
  • Improving the speed and accuracy of diagnosis for genetic diseases.
  • Unlocking the possibility of personalized preventive and medical treatment plans.
  • Regularly monitor patients’ biometric data to see they are complying with their treatment plans.
  • Helping healthcare providers deliver better low-cost, evidence-driven care to consumers.
  • Helping consumers to avoid costly visits to doctor offices and hospitals.
  • Giving everyone in the world the equivalent of a doctor in their pocket – or smartphone.

Future Scenario: By 2040, a space-based global artificial intelligence (AI) network of satellites will be in place that will monitor and help provide healthcare to people on Earth and in colonies across our solar system, e.g. Moon, Mars. The system will be linked to massive global health data warehouses storing data from a wide range of health IT systems, e.g. EHR, PHR, HIE, IoT devices...

The global AI system will monitor and analyze the health data gathered on all humans in real-time, detecting potential individual and public health issues. The system will detect problems, diagnose them, send alerts to patients and their healthcare providers, diagnose the problems and recommend treatment plans to resolve the healthcare issue. The system will also be interfaced to pharmacies, laboratories, health insurers, public health agencies, and other institutions as needed.

Current News & Activities

The following are a few selected articles you might want to read to get a better handle on the latest news about current activities related to Artificial Intelligence (AI) in healthcare:


Selected Issues

The following are some of the key issues that need to be addressed as we continue moving forward with the design, development, and use of AI technologies and data-drive healthcare information systems:

  • Ethics - This emerging issue is concerned with the moral behavior being programmed by humans into robots and other 'smart' machines, i.e. Roboethics.
  • Privacy & SecurityThis is always a key issue. When AI systems are turned loose to monitor all health information systems gathering data on all facets of your personal health, concerns about who has access to the data and who it is being shared with are just a few of the issues that must be adequately addressed upfront.
  • Jobs - The Bank of England has predicted that intelligent machines might take over 80 million American and 15 million British jobs, respectively over the next 10 to 20 years. The healthcare industry will not be immune to this change.
  • Legal Issues - One of the most important points of interest that needs to be hammered out first is the legality of these machines. When a doctor's gross negligence leads to a misdiagnoses and patient harm, the fault is placed squarely on the shoulders of the offending physician. But what happens when a similar situation befalls an AI system? If such a program were to misdiagnose a patient, who's to blame?
  • Open Source - Many new 'open source' tools are arriving that can now run on affordable hardware and allow individuals and small organizations to perform prodigious data crunching and predictive tasks. Read about H2O, OpenAI, and other machine learning and AI tools being used in healthcare at Open Health News
Check out the growing List of AI Projects on Wikipedia.

Regenstrief Institute and Indiana University School of Informatics & Computing, recently examined open source algorithms and machine learning tools in public health reporting: The tools bested human reviewers in detecting cancer using pathology reports and did so faster than people. Indeed, more and more healthcare systems on the cutting edge are looking at ways to harness the power of AI, cognitive computing and machine learning. See Artificial intelligence and cognitive computing - Healthcare IT News

Recommended Next Steps

Artificial Intelligence (AI) systems today can learn in ways society once thought impossible, which has major implications for multiple industries – especially the healthcare industry. It is now time to begin focusing more proactively on a public-private sector collaboration to design and develop a 'global healthcare platform for 2040' built on artificial intelligence (AI) technologies.

Many countries are beginning to support the idea of having a global healthcare network of data centers coupled with a state-of-the-art AI technology platform that will allow sort through all the data, standardize it, and put it into a form that is useful and easily understood by patients, healthcare providers, healthcare insurers, researchers, and other individuals and organizations.

Any such global effort should keep in mind the key management strategies for success in the 21st century – Collaboration, Open Solutions, and continuous Innovation (COSI). Building such a global solution will require a massive global public-private sector partnership. Think of all the components that will need to converge to compose such a global solution – e.g. healthcare technologies, research, knowledge, organizations, and more.

One final note, advances in AI and technology are helping create a futuristic human-to-machine and machine-to-human interaction that can best be described as an 'Invisible' User Interface (IUI) of the future that simply works non-stop in the background to monitor and improve health for everyone. It will just be there – serving mankind.

Recommended Links

Companies
Systems & Projects

News Sites
Associations
Journals


   

3.0 - Invisible User Interface (IUI) for Health 4.0

The Invisible User Interface (IUI) will be a key characteristic of next generation Health 4.0 systems.

The emerging Invisible User Interface (IUI), also referred to as the Natural User Interface (NUI), involves a major paradigm shift in man machine interaction using a computer interface that will be basically invisible. Most computer interfaces today use artificial controls and tangible devices whose operation has to be learned, e.g. Windows, computer mouse, joystick. That's about to change big time with the convergence of multiple modern technologies. See Wikipedia.

The Invisible User Interface (IUI) will include sound, touch, gesture, and tactile inputs and outputs as humans interact with an ever increasing numbers of 'smart' machines all around us, i.e. Internet of Things (IoT). The goal for system developers is to now make ubiquitous computing technology ever more simple to use – designing systems to seemingly be more intuitive and accessible for use by humans with minimal technical knowledge and expertise.

Health 4.0 - Imagine a next generation Artificial Intelligence (AI) system linked to a massive global health data warehouse storing data from a wide range of health IT systems, e.g. Electronic Health Record (EHR) systems, Personal Health Records (PHR), Health Information Exchange (HIE) networks, wearable fitness trackers, implantable medical devices, clinical imaging systems, genomic databases and bio-repositories, health research knowledgebases, medical sensors and more.

Health 4.0 Artificial Intelligence (AI) systems of the future will constantly monitor and analyze the health data gathered on humans in real-time, quietly detecting potential individual and public health issues in the background. It will detect problems and diagnose them, send alerts to patients and their healthcare providers, and generate treatment plans to resolve any healthcare issues. The system will also be interfaced to pharmacies, laboratories, health insurers, public health agencies, and other institutions as needed. The key is that most of this activity will be done quietly in the background.


Health 4.0 systems will use a variety of emerging technologies that will allow an infrastructure of intelligent sensors embedded in our living and working environment to unobtrusively monitor your personal health data and then interact with you, your family, healthcare provider, and other concerned parties using a range advanced verbal and non-verbal communication technologies – think Amazon Echo or Apple's Siri personal assistants.

Invisible User Interface (IUI)

Today's computers can hear, see, read and understand humans better than ever before. Using advanced ambient intelligence devices embedded in the environment where we live and work, these emerging technologies will be able to monitor our movements, voice, glances, and even thoughts – causing these systems to respond and meet our needs in a variety of ways.

These technologies are opening a world of opportunities for AI-powered systems to interact and serve us using next generation Invisible User Interfaces (IUI). Think about it! Our words and natural gestures will trigger interactions with computer systems and the Internet of Things (IoT) all around us, just as if we were communicating to another person. Essentially, we'll be doing away with old-fashioned 'screens', today's graphical user interface (GUI), and the ever growing number of mobile apps.

Building the Invisible User Interface (IUI) is a new way of approaching the user experience that thinks beyond the screen. The goal is to design systems to better fit within our lives, rather than forcing us to adapt to the machines we use. Key features of the IUI and future systems include:

  • Anticipatory design of new products and experiences that use constantly available, real-time data to anticipate what customers need and want to do next.
  • Personalization is the way that companies can better connect and interact with their customers, giving them the information they need in a way that feels more 'human'.
  • Ambient communication involving an infrastructure of intelligent sensors that will be embedded in our living and working environment – including wearable and implantable systems.
  • Artificial Intelligence (AI) will be needed to monitor, analyze, and take action based on the wide array of data being collected on each person.
  • 'Deep learning' is the process of teaching computers to understand and solve a problem by itself, rather than having engineers code each and every solution.

Examples of Systems using Invisible User Interfaces (IUI)

The following are some current examples of emerging next generation apps and interfaces to tomorrow's information systems. What makes them special is that they use a non-traditional invisible or conversational user interfaces as their means of interaction with humans - Read 'No UI' Is The New UI.

  • Amazon EchoVoice interaction device capable of accepting commands and providing a wide range of information to you upon request. It can also control smart devices in your home.
  • Braingate - Creating and testing transformative neurotechnologies to restore communication, mobility, and independence of people with neurologic disease, injury, or limb loss.
  • EmotivCreating mobile EEG wearable systems, providing access to advanced brain monitoring and cognitive assessment technologies. Read Telepathy and Brain Interface
  • FitBit – Creating wearable fitness and health tracking devices for individuals.
  • MagicThis company provides access to a 'smart' personal assistant set of services to meet a range of customer needs.
  • 'M' by Facebook - A personal assistant powered by artificial intelligence (AI) that’s integrated with Messenger to help complete tasks and find information on your behalf.
  • TIII ProjectCollaborative research project focus on developing tangible, intuitive, interactive interfaces (TIII) for the future.
  • WiiUses a handheld pointing device for interacting with the Wii computerized game system that detects movement in three dimensions.
* You might also want to explore Google Glass, Hexoskin, and other 'smart' wearable systems.

Conclusions and Next Steps


How people communicate with each other is very different from how people interact with machines. The trend in computer systems design now involves looking more closely at how humans interact and communicate intuitively. The goal is to teach computers, machines, and the Internet of Things (IoT) to better comprehend and communicate with humans.

The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it.’ That’s what computer scientist Mark Weiser wrote in his famous article The Computer for the 21st Century, originally published in 1991.

In the relatively near future, we will be interacting with computers all around us without noticing them. This ties in closely with the Invisible User Interface (IUI) and the rapid growth of the Internet of Things (IoT), which includes wearable, implantable, and embedded health sensors.

Collaboration, Coherence, and Convergence will be the key to industry efforts to challenge the current generation of health IT and healthcare delivery systems. Efforts are already underway to design and build the next generation healthcare systems – Read Global Health 4.0 for 2040 and Beyond

Other Selected Links





Conclusion - Global collaboration, 'open' solutions, and convergence of the many innovative technologies currently under development are key for plans to build and deploy Health 4.0 by 2040.


* Send us your suggestions and ideas on the development of Health 4.0