This document provides an overview of artificial intelligence in healthcare by an expert in the field. It discusses key concepts like machine learning, deep learning and artificial intelligence. It explores applications of AI in areas like computer vision, vocal biomarkers, and skin cancer detection. It also covers barriers to adoption like trust, transparency and bias issues. Ethical principles from WHO are outlined around human autonomy, safety, transparency and accountability. The impacts on the medical workforce and potential to widen healthcare access gaps are debated. A proposed "Patient AI Bill of Rights" addresses issues like consent, bias, explanations and oversight.

1. THE HEALTHCARE
EXECUTIVE GUIDE
TO INTELLIGENCE-
BASED MEDICINE
ARLEN MEYERS, MD, MBA
EMERITUS PROFESSOR, UNIVERSITY OF COLORADO
SCHOOL OF MEDICINE
PRESIDENT AND CEO , SOCIETY OF PHYSICIAN
ENTREPRENEURS
ADVIDSOR AND PRINCIPAL, MI10

2. HYPE OR
HOPE?

6. WHAT YOU
WILL
LEARN
The basic concepts of artificial intelligence in medicine
The barriers to dissemination and implementation
The hazards of using AI in medicine
Learn how AI is being used in medicine to improve quality,
access, your experience, doctor experience and business
processes at lower costs.

7. PLEASE DISCUSS
Google maps, Alexa,
customer service chat bots,
some mobile medical apps
and Smart TVs use artificial
intelligence.
What has been your
experience using these?
Concerns?

8. WHAT IS HEALTHCARE ENTREPRENEURSHIP?
Pursuit of opportunity
Under VUCA (volatile,
uncertain, complex,
ambiguous)
conditions
Goal is to create
stakeholder defined
value
Through the
deployment of
innovation
Using a VAST (viable,
austomatic, scaleable,
time sensitive)
businsess model

9. WHAT IS DIGITAL HEALTH?
THE APPLICATION OF
INFORMATION AND
COMMUNICATION
TECHNOLOGIES (ICTS) TO
EXCHANGE MEDICAL
INFORMATION FOR VARIOUS
INTENDED USES
• Telehealth
• Electronic medical records
• Big data and analytics
• Remote patient monitoring
• Patient reported outcomes
Virtual and augmented reality
Blockchain
Artificial intelligence
Mobile medical apps
Digital therapeutics

10. WHAT IS ARTIFICIAL INTELLIGENCE?
• There is no universal definition of artificial intelligence (AI). AI is generally
considered to be a discipline of computer science that is aimed at developing
machines and systems that can carry out tasks considered to require human
intelligence. Machine learning and deep learning are two subsets of AI. In
recent years, with the development of new neural networks techniques and
hardware, AI is usually perceived as a synonym for “deep supervised
machine learning”.

11. WHAT IS MACHINE LEARNING
• Machine learning uses examples of input and expected output (so called
“structured data” or “training data”), in order to continually improve and make
decisions without being programmed how to do so in a step-by-step
sequence of instructions. This approach mimics actual biological cognition: a
child learns to recognize objects (such as cups) from examples of the same
objects (such as various kinds of cups). Today application of machine
learning are widespread including email spam filtering, machine translation,
voice, text and image recognition.

12. WHAT IS DEEP LEARNING?
• Deep learning has evolved from machine learning. Deep learning uses a
plurality of AI algorithms (so called “artificial neural networks”) to recognize
patterns, hence being able to group and classify unlabeled data.

14. APPLICATIONS
• Future clinical evidence generation with real
world data
• Patient-centered care with wearable
technology with embedded AI
• Advanced automated processes with robotic
process automation
• Continual learning health system with edge
computing
• Widespread citizen medicine with artificial
intelligence tools
• Innovative electronic health record with graph
database
• Real-time decision support with deep
reinforcement learning
• Intelligent reality with AI and virtual reality
• Future clinical research with AI and virtual
twins
• Advanced communications with transformer
type natural language processing

15. BARRIERS TO AI DISSEMINATION AND
IMPLEMENTATION
There are four basic
categories of barriers:
1) technical 2) human factors
3) environmental, including
legal, regulatory, ethical,
political, societal and
economic determinants and
4) business model barriers
to entry.

16. HAZARDS
AND
LANDMINE
S
•Trust
•The Black Box problem:
transparency
•Bias
•Regulatory, economic,
ethical and societal
issues

18. TRUST STANDARDS
• Known as ANSI/CTA-2090, "The Use of Artificial Intelligence in Health
Care: Trustworthiness" – considers what the association says are the
three key areas relating to how trust is created and maintained across
stakeholders.

19. HUMAN TRUST
• Human trust is concerned with developing "humanistic factors that
affect the creation and maintenance of trust between the developer
and users," according to CTA. "Specifically, human trust is built upon
human interaction, the ability to easily explain, user experience and
levels of autonomy of the AI solution."

20. TECHNICAL TRUST
• Technical trust is focused on design and training of AI and machine
learning systems, ensuring they "deliver results as expected."
Additionally, it considers data quality and integrity – such as
algorithmic bias, security, privacy, source and access.

21. REGULATORY TRUST
• Regulatory trust, meanwhile, is "gained through compliance by
industry based upon clear laws and regulations," said CTA, whether
that's from accreditation boards, regulatory agencies, federal and
state laws or international standardization frameworks.

23. BUT, WHAT ABOUT CYBERSECURITY?

24. PLEASE DISCUSS
What are your concerns
about the ethical use of AI
in medicine?
Suppose you were treated
by a doctor who used AI to
make a diagnosis which
turned out to be wrong.
What would you do?

25. WHO ETHICAL PRINCIPLES
Protecting human autonomy: In the context of health care, this means
that humans should remain in control of healthcare systems and
medical decisions; privacy and confidentiality should be protected, and
patients must give valid informed consent through appropriate legal
frameworks for data protection.

26. WHO PRINCIPLES
• Promoting human well-being and safety and the public
interest. The designers of AI technologies should satisfy regulatory
requirements for safety, accuracy and efficacy for well-defined use
cases or indications. Measures of quality control in practice and
quality improvement in the use of AI must be available.

27. WHO PRINCIPLES
• Ensuring transparency, explainability and
intelligibility. Transparency requires that sufficient information be
published or documented before the design or deployment of an AI
technology. Such information must be easily accessible and facilitate
meaningful public consultation and debate on how the technology is
designed and how it should or should not be used.

28. WHO PRINCIPLES
• Fostering responsibility and accountability. Although AI
technologies perform specific tasks, it is the responsibility of
stakeholders to ensure that they are used under appropriate
conditions and by appropriately trained people. Effective mechanisms
should be available for questioning and for redress for individuals and
groups that are adversely affected by decisions based on algorithms.

29. WHO PRINCIPLES
• Ensuring inclusiveness and equity. Inclusiveness requires that AI
for health be designed to encourage the widest possible equitable use
and access, irrespective of age, sex, gender, income, race, ethnicity,
sexual orientation, ability or other characteristics protected under
human rights codes.

30. WHO PRINCIPLES
• Promoting AI that is responsive and sustainable. Designers, developers
and users should continuously and transparently assess AI applications
during actual use to determine whether AI responds adequately and
appropriately to expectations and requirements. AI systems should also be
designed to minimize their environmental consequences and increase
energy efficiency. Governments and companies should address anticipated
disruptions in the workplace, including training for health-care workers to
adapt to the use of AI systems, and potential job losses due to use of
automated systems.

31. USE CASES IN MEDICINE
Radiology and imaging Clinical decision support Personalized medicine Robotic process
automation
Patient education,
experience, engagement
and behavior change
enablement

32. AI CHALLENGES IN HEALTHCARE
• Addressing the ethical, legal, and societal implications of AI
• Ensuring the safety and security of AI systems
• Developing shared public datasets and environments for AI training and
testing
• Evaluating AI technologies using standards and benchmarks
• Understanding the national AI R&D workforce needs
• Expanding public-private partnerships to accelerate advances in AI

33. DIGITAL TRANSFORMATION AND
DEXTERITY
• CULTURE
• PROCESSES
• TECHNOLOGY
• PEOPLE
• LEADERSHIP
• ALIGNMENT
• EDUCATION AND TRAINING

35. 10 COMPUTER VISION APPLICATIONS
APPLICATIONS
• Blood loss measurement
• Cardiology ultrasound/ekg
• Tumor detection
• Cancer detection
• Illness prediction
APPLICATIONS
• AR/VR in surgery
• COVID staff. Stuff and space
• Voice diagnosis
• Retinal scanning
• Melanoma detection

36. VOCALISCHECK
• AI-based vocal biomarker company Vocalis Health received a CE
mark for its Covid-19 'voice check' tool. VocalisCheck is a software-
only product that collects a single voice sample from users simply by
counting from 50 to 70. The recording is transformed to a picture
(spectrogram) containing 512 features which are analysed by AI
machine learning. The tool was recently validated in a large clinical
study demonstrating an accuracy sensitivity and specificity above
80%, even among asymptomatic cohorts

37. IMAGE DETECTION OF SKIN CANCER
• AI can be taught to flag possible skin cancers on photos taken with
smartphone cameras—and the images can be ordinary “people shots”
rather than closeups of suspicious lesions. Using lesion classification
by experienced dermatologists as the ground truth, the researchers
found their system achieved sensitivity and specificity of right around
90% each when tasked with separating suspicious lesions from
benign skin discolorations and busy backgrounds.

38. USING METADATA
• https://www.youtube.com/watch?v=
LH2F8PNJdu0
What do I do?

39. PLEASE DISCUSS
• Will the use of artificial intelligence in medicine eliminate
doctors or other healthcare professionals?
• Why or why not? Would you want it to?
• How will it impact the future of medical education and
workforce needs?
• Will AI widen the gaps between the haves and have-nots?

40. PATIENT AI BILL OF RIGHTS
• Transparency: We have the right to know when an algorithm is making a
decision about us, which factors are being considered by the algorithm, and
how those factors are being weighted.
• Explanation: We have the right to be given explanations about how
algorithms affect us in a specific situation, and these explanations should be
clear enough that the average person will be able to understand them.
• https://www.vox.com/the-highlight/2019/5/22/18273284/ai-
algorithmic-bill-of-rights-accountability-transparency-consent-bias

41. BILL OF RIGHTS
• Consent: We have the right to give or refuse consent for any AI
application that has a material impact on our lives or uses sensitive
data, such as biometric data.
• Freedom from bias: We have the right to evidence showing that
algorithms have been tested for bias related to race, gender, and
other protected characteristics — before they’re rolled out. The
algorithms must meet standards of fairness and nondiscrimination and
ensure just outcomes.

42. BILL OF RIGHTS
• Feedback mechanism: We have the right to exert some degree of
control over the way algorithms work.
• Portability: We have the right to easily transfer all our data from one
provider to another.
• Redress: We have the right to seek redress if we believe an
algorithmic system has unfairly penalized or harmed us.

43. BILL OF RIGHTS
• Algorithmic literacy: We have the right to free educational resources about
algorithmic systems.
• Independent oversight: We have the right to expect that an independent
oversight body will be appointed to conduct retrospective reviews of
algorithmic systems gone wrong. The results of these investigations should
be made public.
• Federal and global governance: We have the right to robust federal and
global governance structures with human rights at their center. Algorithmic
systems don’t stop at national borders, and they are increasingly used to
decide who gets to cross borders, making international governance crucial.

44. 7 LESSONS
• #1 AI investment has weathered Covid disruption
• #2 AI is in a process of democratization
• #3 New initiatives have high-level endorsement from C-Suite
• #4 Customer service will remain one of the key applications
• #5 Some use AI use cases are finding homes on other hype cycles
• #6 Gartner recommends a focus on narrow use cases
• #7 The AI hype cycle makes it impossible to pick winners or losers
• https://www.babelforce.com/blog/perspectives/7-lessons-from-gartners-ai-hype-cycle-2020/

45. WANT TO LEARN MORE?
• Free online introduction to artificial intelligence
• https://www.elementsofai.com/
• www.abaim.org
• www.misociety.org
• www.MI10.ai
• www.AI-med.io

46. THANK YOU