by Vaikunthan Rajaratnam
MBBS,FRCS(Ed),MBA,MIDT, PhD(Education).
Senior Consultant, Hand & Reconstructive Microsurgery, KTPH, Singapore
Program Director Instructional Design for Healthcare & AI in Healthcare NHG, Singapore
Adjunct Professor/UNESCO Chair Partner, Asia Pacific University of Technology and Innovation, Malaysia
Honorary Professor, De Montfort University, UK
Embracing Change,
Unlearning, and Relearning:
Adapting to the Future of Healthcare
Warning:
Unprecedented Levels of Productivity
and Inspiration Ahead!
Please be advised:
The content of this workshop is so intensely engaging and
empowering in the realm of AI in healthcare that it
carries a high risk of sparking a newfound addiction to
productivity and innovation. Attendees may experience
an irresistible urge to apply transformative skills and
insights in their professional practice, leading to
significant advancements in healthcare. Proceed with
enthusiasm and caution – you're stepping into a world of
exhilarating empowerment!
Embrace the journey, but don't say we didn't warn you!
Disclaimer
• I am not an AI expert, nor do I possess coding knowledge specific to the underlying mechanisms of AI models;
• My expertise lies in the utilisation of these models, such as ChatGPT, based on my extensive experience as a
user within the fields of healthcare, medical education, and related research, rather than their technical
development or underlying algorithms.
• This workshop is intended solely for educational and informational purposes in AI and healthcare.
• The views expressed herein are my own, borne from extensive experience in surgery, medical education, and
instructional design, and do not necessarily reflect those of any associated institutions.
• While I endeavour to provide accurate and up-to-date information, no guarantee is given regarding its applicability.
• Participants acknowledge and assume responsibility for using the information provided by engaging in this
workshop.
Vaikunthan Rajaratnam
Innovative Digital Learning
Centre
Artificial Intelligence Innovative Designs
for Education (AIDE) Framework
• Tools for Effective Academic Courses and Holistic Teaching
• MOE Malaysia
AI TEACH
• For students in middle school
• Indigenous School, Malaysia
AI LEARN
• Gerontological Optimization through Learning and Digital Assistance
AI GOLD
• Learning Designs
• APU, Malaysia, NHG, Singapore
AI LD
• Health Professional
• Perdana University , Malaysia, BDSSH, Bangladesh
• NHG, Singapore, Sengkang, Singapore
AI HP
• Research
• Perdana University, Malaysia, Sengkang , Singapore
• University of Eswatini (Africa)
AI RE
•Academic Writing
• APU, Perdana University, Malaysia
• University of Eswatini (Africa)
AI AW
• Leveraging Efficiency in Administrative Proficiency
• MOE, Dubai
AI LEAP
Main Findings on
Unlearning and
Relearning in
Healthcare
A Summary of Key
Research Insights
Unlearning Outdated
Practices: Embracing a
Mindset of Continuous
Learning and Improvement
1 Challenge Assumptions
Regularly question the status quo and be open to new approaches that
challenge conventional wisdom.
2 Foster a Culture of Innovation
Encourage a mindset of continuous learning and experimentation to drive
meaningful change.
3 Adapt to Evolving Needs
Stay agile and responsive to the changing demands of patients and the
healthcare landscape.
Physician Unlearning and Clinical Practice Changes
• Physicians face challenges in
unlearning outdated practices.
• Incomplete unlearning -
increased medical errors.
• Structured unlearning
processes are needed to prevent
patient harm.
(Noteboom, Hafner, & Wahbeh, 2017; Hafner, 2016)
Organisational Unlearning
Unlearning is
critical for adapting
to new practices
and technologies.
Effective unlearning
improves patient
safety and care
outcomes.
(Sharma & Lenka, 2019)
Link Between Unlearning and Innovation
Unlearning
is essential
for
fostering
innovation
in
healthcare.
Relearning
fails
without
deliberate
unlearning
of obsolete
practices.
(Bedford, 2015)
Impact on Patient Safety
Failure to unlearn
outdated practices -
patient safety issues.
Focusing on
unlearning can
reduce medical
errors and improve
care quality.
(Gupta, Boland, & Aron, 2017)
Policy
Recommendations for
Unlearning in
Healthcare
Structured Approaches to Enhance Professional
Practice and Patient Safety
This Photo by Unknown Author is licensed under CC BY
Structured Unlearning Processes
• Develop structured
processes for unlearning
outdated practices.
• Ensure smooth
transition to new
protocols and guidelines.
• Focus on patient safety
and reducing medical
errors.
Rationale for Structured Unlearning
Stay current with
clinical
guidelines.
Reduces
inefficiencies
and improves
resource use.
Patient safety
by eliminating
outdated
practices.
Overview of
Structured
Unlearning
Unlearning involves
intentionally discarding
outdated knowledge and
practices.
Vital for adopting new,
evidence-based practices in
healthcare.
Helps prevent patient harm Improves efficiency.
• Conduct regular assessments and audits.
• Implement feedback systems for reporting outdated practices.
Identify Outdated
Practices
• Develop educational programs and communication strategies.
• Apply change management principles to support acceptance.
Create Awareness
• Establish structured unlearning programs within professional development.
• Integrate unlearning into medical education curricula.
Develop Unlearning
Protocols
• Organize workshops, simulations, and training sessions.
• Set up mentorship and peer learning programs.
Implement Training
and Support
• Continuously reinforce new practices through practice and assessments.
• Monitor and evaluate the effectiveness of unlearning initiatives.
Reinforce New
Learning
• Integrate unlearning into organisational policies and procedures.
• Foster a culture of continuous improvement and open dialogue.
Institutionalise
Unlearning
• Regularly review and update unlearning processes.
• Establish a feedback loop for continuous improvement
Evaluate and Update
Integrate
unlearning into
medical education
and professional
development.
Establish
mentorship and
peer learning
programs.
Conduct regular
audits and provide
feedback on
clinical practices.
Implementation Strategies
Resistance to change
among experienced
professionals.
Unlearning should be
seen as an ongoing
process, not a one-
time event.
Requires careful change
management and
involvement of staff.
Challenges and
Considerations
Embed
unlearning as a
core component
in medical
curricula.
Prepare future
healthcare
professionals to
adapt to change.
Improve the
transition to
new clinical
guidelines and
technologies.
Integration in Medical Education
Support Systems for Unlearning
Establish mentorship and peer support
programs.
Provide regular assessments and
feedback to ensure effective unlearning.
Foster a culture of continuous
improvement and adaptation.
Patient Safety Reforms
Integrate
unlearning
strategies into
patient safety
policies.
Mitigate risks
associated
with outdated
practices.
Focus on
reducing
errors and
enhancing
care quality.
Revolutionising
Healthcare through AI
AI technologies have the potential to
revolutionise healthcare by enhancing
diagnosis, treatment planning, and research.
AI won't replace you, but
someone empowered by AI
undoubtedly will
Upskilling Healthcare
Professionals: Adapting to the
AI-Augmented Future
Digital Literacy
Develop proficiency in leveraging AI-driven tools and technologies to
enhance patient care.
Data Analytics
Cultivate skills in interpreting and applying data-driven insights generated
by generative AI.
Ethical AI Practices
Understand and implement ethical guidelines for the responsible use of AI
in healthcare.
Relearning Through Data-Driven
Insights: Leveraging Generative AI for
Better Decision-Making
Predictive Analytics
Utilize generative AI to identify
patterns and trends, enabling
proactive and personalized care.
Clinical Decision Support
Empower clinicians with AI-
powered recommendations and
decision-making tools for
improved patient outcomes.
Streamlined Workflows
Integrate generative AI into daily
tasks to automate repetitive
processes and enhance
efficiency.
Integrating Generative AI
into Daily Clinical Workflows
1 Data Collection
Leverage AI-powered sensors and devices to gather
comprehensive patient data.
2 Intelligent Insights
Utilize generative AI to analyze data and provide personalized
recommendations.
3 Streamlined Workflows
Integrate AI-driven tools into daily clinical processes for enhanced
efficiency.
Enhancing Patient Engagement
and Communication with AI-
Powered Tools
Conversational AI
Intelligent chatbots and virtual assistants for seamless patient communication.
Personalized Guidance
AI-powered tools that provide tailored health advice and support for patients.
Increased Engagement
AI-driven platforms that foster stronger patient-provider relationships.
Ethical Considerations and Responsible AI
Implementation
Transparency and Accountability
Ensure clear communication and ethical guidelines
for the use of generative AI in healthcare.
Patient Privacy and Data Security
Prioritize the protection of sensitive patient
information and adhere to data privacy regulations.
Bias Mitigation
Proactively address and minimize biases in AI-driven
decision-making and patient care.
Continuous Monitoring
Establish robust monitoring and evaluation
processes to ensure responsible AI implementation.
Introduction to Generative AI
https://chat.openai.com/ https://copilot.microsoft.com/ https://bard.google.com/
Generative AI Platforms
Opening a ChatGPT Account
chatgpt.com
ChatGPT for Android
Prompt
Generation Steps
Define Your
Objective
•Be specific
•Overall goal
Understand
Your AI Model
•Model capabilities
•Data Reliability
Break Down
Complex Tasks
•Deconstruct
•Sequence
Craft Clear and
Concise
Prompts
•Avoid ambiguity
•Provide context
•Specify outputs
Start with
Examples (Few-
Shot Learning)
•Illustrative
input/output
Iterate and
Refine
•Analyse responses
•Adjust language
•Test, test, test
Checklist for Healthcare Prompt Engineering
Response Validation
RI meets AI
• Review response - meets your requirements.
• No access to real-time data
• Vaildate Validate Validate.
• Prompt – response -refine - reprompt.
Relevance Check
Accuracy
Confirmation
Context
Consistency
Sensitivity Review
Refinement for
Future Queries
Review AI Output with
RI
•Critically assess
•Relevance
•Plausibility
Key Term Extraction
•Extract central terms
Strategic Search in
Academic Databases
•Focused search
•Using Extracted terms
•Employing advanced search
techniques
Literature Screening
•Select relevant scholarly articles
•Prioritising high-quality, peer-
reviewed sources
•In-depth analysis.
Evidence-Based
Validation
•Critically evaluate the literature
•Corroborate or challenge the AI-
generated information
•Look for consistency and
consensus
Synthesis and
Documentation
•Compile and synthesise findings
•Validate the AI’s responses
•Document and cite the sources
Evidence-based
Validation
http://tinyurl.com/AIVIDEOUP
Dictate
Insert prompt
Upload to
MyGPT
Send EMR to
email
Review & Edit
Insert to
Institution
EMR
Intelligent Tutoring Systems
Patient
Education
Support
Systems
https://tinyurl.com/AIHPOER24
https://tinyurl.com/AIHPVIDEOS
Contact
vaikunthan@gmail.com

WHC Keynote Learn^JUnlearn and Relearn Sept 2024.pdf

  • 1.
    by Vaikunthan Rajaratnam MBBS,FRCS(Ed),MBA,MIDT,PhD(Education). Senior Consultant, Hand & Reconstructive Microsurgery, KTPH, Singapore Program Director Instructional Design for Healthcare & AI in Healthcare NHG, Singapore Adjunct Professor/UNESCO Chair Partner, Asia Pacific University of Technology and Innovation, Malaysia Honorary Professor, De Montfort University, UK Embracing Change, Unlearning, and Relearning: Adapting to the Future of Healthcare
  • 2.
    Warning: Unprecedented Levels ofProductivity and Inspiration Ahead! Please be advised: The content of this workshop is so intensely engaging and empowering in the realm of AI in healthcare that it carries a high risk of sparking a newfound addiction to productivity and innovation. Attendees may experience an irresistible urge to apply transformative skills and insights in their professional practice, leading to significant advancements in healthcare. Proceed with enthusiasm and caution – you're stepping into a world of exhilarating empowerment! Embrace the journey, but don't say we didn't warn you!
  • 3.
    Disclaimer • I amnot an AI expert, nor do I possess coding knowledge specific to the underlying mechanisms of AI models; • My expertise lies in the utilisation of these models, such as ChatGPT, based on my extensive experience as a user within the fields of healthcare, medical education, and related research, rather than their technical development or underlying algorithms. • This workshop is intended solely for educational and informational purposes in AI and healthcare. • The views expressed herein are my own, borne from extensive experience in surgery, medical education, and instructional design, and do not necessarily reflect those of any associated institutions. • While I endeavour to provide accurate and up-to-date information, no guarantee is given regarding its applicability. • Participants acknowledge and assume responsibility for using the information provided by engaging in this workshop. Vaikunthan Rajaratnam
  • 4.
    Innovative Digital Learning Centre ArtificialIntelligence Innovative Designs for Education (AIDE) Framework
  • 6.
    • Tools forEffective Academic Courses and Holistic Teaching • MOE Malaysia AI TEACH • For students in middle school • Indigenous School, Malaysia AI LEARN • Gerontological Optimization through Learning and Digital Assistance AI GOLD • Learning Designs • APU, Malaysia, NHG, Singapore AI LD • Health Professional • Perdana University , Malaysia, BDSSH, Bangladesh • NHG, Singapore, Sengkang, Singapore AI HP • Research • Perdana University, Malaysia, Sengkang , Singapore • University of Eswatini (Africa) AI RE •Academic Writing • APU, Perdana University, Malaysia • University of Eswatini (Africa) AI AW • Leveraging Efficiency in Administrative Proficiency • MOE, Dubai AI LEAP
  • 7.
    Main Findings on Unlearningand Relearning in Healthcare A Summary of Key Research Insights
  • 8.
    Unlearning Outdated Practices: Embracinga Mindset of Continuous Learning and Improvement 1 Challenge Assumptions Regularly question the status quo and be open to new approaches that challenge conventional wisdom. 2 Foster a Culture of Innovation Encourage a mindset of continuous learning and experimentation to drive meaningful change. 3 Adapt to Evolving Needs Stay agile and responsive to the changing demands of patients and the healthcare landscape.
  • 9.
    Physician Unlearning andClinical Practice Changes • Physicians face challenges in unlearning outdated practices. • Incomplete unlearning - increased medical errors. • Structured unlearning processes are needed to prevent patient harm. (Noteboom, Hafner, & Wahbeh, 2017; Hafner, 2016)
  • 10.
    Organisational Unlearning Unlearning is criticalfor adapting to new practices and technologies. Effective unlearning improves patient safety and care outcomes. (Sharma & Lenka, 2019)
  • 11.
    Link Between Unlearningand Innovation Unlearning is essential for fostering innovation in healthcare. Relearning fails without deliberate unlearning of obsolete practices. (Bedford, 2015)
  • 12.
    Impact on PatientSafety Failure to unlearn outdated practices - patient safety issues. Focusing on unlearning can reduce medical errors and improve care quality. (Gupta, Boland, & Aron, 2017)
  • 13.
    Policy Recommendations for Unlearning in Healthcare StructuredApproaches to Enhance Professional Practice and Patient Safety This Photo by Unknown Author is licensed under CC BY
  • 14.
    Structured Unlearning Processes •Develop structured processes for unlearning outdated practices. • Ensure smooth transition to new protocols and guidelines. • Focus on patient safety and reducing medical errors.
  • 15.
    Rationale for StructuredUnlearning Stay current with clinical guidelines. Reduces inefficiencies and improves resource use. Patient safety by eliminating outdated practices.
  • 16.
    Overview of Structured Unlearning Unlearning involves intentionallydiscarding outdated knowledge and practices. Vital for adopting new, evidence-based practices in healthcare. Helps prevent patient harm Improves efficiency.
  • 17.
    • Conduct regularassessments and audits. • Implement feedback systems for reporting outdated practices. Identify Outdated Practices • Develop educational programs and communication strategies. • Apply change management principles to support acceptance. Create Awareness • Establish structured unlearning programs within professional development. • Integrate unlearning into medical education curricula. Develop Unlearning Protocols • Organize workshops, simulations, and training sessions. • Set up mentorship and peer learning programs. Implement Training and Support • Continuously reinforce new practices through practice and assessments. • Monitor and evaluate the effectiveness of unlearning initiatives. Reinforce New Learning • Integrate unlearning into organisational policies and procedures. • Foster a culture of continuous improvement and open dialogue. Institutionalise Unlearning • Regularly review and update unlearning processes. • Establish a feedback loop for continuous improvement Evaluate and Update
  • 18.
    Integrate unlearning into medical education andprofessional development. Establish mentorship and peer learning programs. Conduct regular audits and provide feedback on clinical practices. Implementation Strategies
  • 19.
    Resistance to change amongexperienced professionals. Unlearning should be seen as an ongoing process, not a one- time event. Requires careful change management and involvement of staff. Challenges and Considerations
  • 20.
    Embed unlearning as a corecomponent in medical curricula. Prepare future healthcare professionals to adapt to change. Improve the transition to new clinical guidelines and technologies. Integration in Medical Education
  • 21.
    Support Systems forUnlearning Establish mentorship and peer support programs. Provide regular assessments and feedback to ensure effective unlearning. Foster a culture of continuous improvement and adaptation.
  • 22.
    Patient Safety Reforms Integrate unlearning strategiesinto patient safety policies. Mitigate risks associated with outdated practices. Focus on reducing errors and enhancing care quality.
  • 23.
    Revolutionising Healthcare through AI AItechnologies have the potential to revolutionise healthcare by enhancing diagnosis, treatment planning, and research. AI won't replace you, but someone empowered by AI undoubtedly will
  • 24.
    Upskilling Healthcare Professionals: Adaptingto the AI-Augmented Future Digital Literacy Develop proficiency in leveraging AI-driven tools and technologies to enhance patient care. Data Analytics Cultivate skills in interpreting and applying data-driven insights generated by generative AI. Ethical AI Practices Understand and implement ethical guidelines for the responsible use of AI in healthcare.
  • 25.
    Relearning Through Data-Driven Insights:Leveraging Generative AI for Better Decision-Making Predictive Analytics Utilize generative AI to identify patterns and trends, enabling proactive and personalized care. Clinical Decision Support Empower clinicians with AI- powered recommendations and decision-making tools for improved patient outcomes. Streamlined Workflows Integrate generative AI into daily tasks to automate repetitive processes and enhance efficiency.
  • 26.
    Integrating Generative AI intoDaily Clinical Workflows 1 Data Collection Leverage AI-powered sensors and devices to gather comprehensive patient data. 2 Intelligent Insights Utilize generative AI to analyze data and provide personalized recommendations. 3 Streamlined Workflows Integrate AI-driven tools into daily clinical processes for enhanced efficiency.
  • 27.
    Enhancing Patient Engagement andCommunication with AI- Powered Tools Conversational AI Intelligent chatbots and virtual assistants for seamless patient communication. Personalized Guidance AI-powered tools that provide tailored health advice and support for patients. Increased Engagement AI-driven platforms that foster stronger patient-provider relationships.
  • 28.
    Ethical Considerations andResponsible AI Implementation Transparency and Accountability Ensure clear communication and ethical guidelines for the use of generative AI in healthcare. Patient Privacy and Data Security Prioritize the protection of sensitive patient information and adhere to data privacy regulations. Bias Mitigation Proactively address and minimize biases in AI-driven decision-making and patient care. Continuous Monitoring Establish robust monitoring and evaluation processes to ensure responsible AI implementation.
  • 29.
  • 30.
  • 31.
    Opening a ChatGPTAccount chatgpt.com
  • 33.
  • 34.
    Prompt Generation Steps Define Your Objective •Bespecific •Overall goal Understand Your AI Model •Model capabilities •Data Reliability Break Down Complex Tasks •Deconstruct •Sequence Craft Clear and Concise Prompts •Avoid ambiguity •Provide context •Specify outputs Start with Examples (Few- Shot Learning) •Illustrative input/output Iterate and Refine •Analyse responses •Adjust language •Test, test, test
  • 35.
    Checklist for HealthcarePrompt Engineering
  • 36.
    Response Validation RI meetsAI • Review response - meets your requirements. • No access to real-time data • Vaildate Validate Validate. • Prompt – response -refine - reprompt. Relevance Check Accuracy Confirmation Context Consistency Sensitivity Review Refinement for Future Queries
  • 37.
    Review AI Outputwith RI •Critically assess •Relevance •Plausibility Key Term Extraction •Extract central terms Strategic Search in Academic Databases •Focused search •Using Extracted terms •Employing advanced search techniques Literature Screening •Select relevant scholarly articles •Prioritising high-quality, peer- reviewed sources •In-depth analysis. Evidence-Based Validation •Critically evaluate the literature •Corroborate or challenge the AI- generated information •Look for consistency and consensus Synthesis and Documentation •Compile and synthesise findings •Validate the AI’s responses •Document and cite the sources Evidence-based Validation
  • 42.
  • 46.
    Dictate Insert prompt Upload to MyGPT SendEMR to email Review & Edit Insert to Institution EMR
  • 48.
  • 51.
  • 53.
  • 54.
  • 55.