Leveraging AI and Dashboards for Public Health Surveillance and Reporting

Kamarul Imran Musa (KIM)

Profile

Contents

  • Dashboards in public health
  • Core functions in disease surveillance
  • Predictive and Generative AI
  • Benefits of integrating AI in dashboards
  • Infra and software requirements
  • Knowledge and skills to build AI-Powered dashboards

Dashboard Concepts in Public Health

What is a Public Health Dashboard?

Definition: An intelligence tool that uses data visualization and analytics to provide actionable feedback for:

  • Disease surveillance
  • Outbreak detection
  • Resource allocation
  • Performance monitoring
  • Evidence-based decision making

Key Components:

  • Data integration
  • Real-time updates
  • Interactive visualizations
  • User-friendly interface
  • Multi-level access

“Dashboards allow users to quickly visualize actionable data to inform and optimize clinical and organizational performance”

John Hopkins COVID-19 Dashboard

Latest John Hopkins COVID-19 Dashboard

John Hopkins COVID-19

Johns Hopkins COVID-19 Dashboard Development

Origin and Timeline

Launch Details:

  • Date: January 22, 2020
  • Institution: Johns Hopkins University
  • Purpose: First global real-time coronavirus surveillance
  • Team: 8-10 full-time equivalents
  • Technology: ArcGIS Dashboard platform

Initial Scope:

  • Cases tracking
  • Deaths monitoring
  • Recovery data
  • Geographic mapping
  • Time series analysis

“Milestone: Became the de facto global database for SARS-CoV-2 spread tracking”

Other examples

Institute of Health Metrics and Evaluation (IHME)

Institute of Health Metrics and Evaluation

Planetary Child Health and Enterics Observatory

Integrating Predictive AI in Dashboards

Machine Learning Applications

Surveillance Applications:

  • Outbreak detection algorithms
  • Disease spread modeling
  • Contact tracing optimization
  • Resource allocation prediction
  • Risk stratification models

Implementation Examples:

  • SIR-DL model for COVID-19
  • HealthMap surveillance system
  • Enhanced Visual Assessment (EVA) border screening system
  • Enhanced Detection System for Healthcare-Associated Transmission (EDS-HAT) hospital surveillance
  • Influenza forecasting models

“Making prediction to assist in decision making, triaging, making diagnosis, identifying high risk individuals and generating spatial prediction”

Integrating Generative AI in Dashboards

Text Generation Applications

Content Creation:

  • Automated report generation
  • Clinical documentation
  • Policy summaries
  • Research abstracts
  • Educational materials

Communication Tools:

  • Public health messaging
  • Risk communication
  • Multilingual translation
  • Plain language summaries
  • Interactive explanations

“Generating new objects. Objects could be guidelines, policies, new records, summary, explain laboratory results and reports.”

Image and Visualization Generation

Visual Content:

  • Infographic creation
  • Chart generation
  • Map visualizations
  • Interactive graphics
  • Educational diagrams

Applications:

  • Public awareness campaigns
  • Training materials
  • Data storytelling
  • Risk visualization
  • Scenario modeling

“Advantage: Enables rapid creation of compelling, accessible visual and image communications.”

Example

Integrating Both AIs in Dasboards

Combined AI Capabilities

graph LR
A[Data Input] --> B[Predictive AI]
A --> C[Generative AI]
B --> D[Risk Scores]
B --> E[Forecasts]
B --> F[Classifications]
C --> G[Reports]
C --> H[Recommendations]
C --> I[Summaries]
D --> J[Dashboard Integration]
E --> J
F --> J
G --> J
H --> J
I --> J

Software and Infrastructure Requirements

Technology Stack

Development Platforms:

  • R/RStudio ecosystem
  • Python environments
  • Cloud computing platforms
  • Containerization (Docker)
  • Version control (Git)

AI Frameworks:

  • TensorFlow/Keras
  • PyTorch
  • Scikit-learn
  • Transformers library
  • MLflow for model management

Infrastructure Components

Computing Resources:

  • CPU clusters
  • GPU acceleration
  • Cloud services (AWS, Azure, GCP)
  • Scalable storage systems

Deployment Options:

  • Shinyapps.io
  • Posit Connect
  • Kubernetes clusters
  • Hybrid cloud solutions

“Considerations: Balance performance, cost, security, and scalability requirements.”

Programming Knowledge for AI-Powered Dashboards

R Ecosystem Mastery

Core R Skills:

  • Data manipulation (dplyr, tidyr)
  • Statistical modeling
  • Time series analysis
  • Geospatial analysis
  • API integration

Shiny Development:

  • Reactive programming
  • UI/UX design principles
  • Module development
  • Performance optimization
  • User session management

“Of course, this slide applies if you a R hardcore fan like me but, i do love and use Python, sometimes”

Programming Skills for AI-Powered Dashboards

Technical Competency Framework

Foundation Skills:

  • Statistical programming in R
  • Database management
  • API development and consumption
  • Version control with Git
  • Reproducible research practices

Advanced Capabilities:

  • Machine learning model development
  • Real-time data processing
  • Scalable application architecture
  • Security implementation
  • Performance optimization

Conclusion

Transformative Impact

Dashboard Evolution:

  • Static reporting → Interactive intelligence
  • Manual processes → Automated insights
  • Reactive responses → Proactive interventions
  • Isolated systems → Integrated ecosystems

AI Enhancement:

  • Pattern recognition at scale
  • Predictive capabilities
  • Natural language interaction
  • Data interpretation
  • Generative content creation (suggestion, actions, contents)
  • Intelligent automation (Agent AI or Agentic AI)

Key Takeaways

Strategic Imperatives

  1. Invest in AI-powered surveillance infrastructure
  2. Develop technical capabilities in R and Shiny ecosystem
  3. Establish data quality, sharing and governance frameworks
  4. Foster regional and global collaboration
  5. Prioritize ethical AI implementation
  6. Redesign undergraduate and postgraduate training in data analysis, modelling and analytics

Success Factors

  • Support and commitment from leadership
  • Local domain expertise in public health
  • Technical proficiency in modern AI tools
  • Understanding of AI-powered deployment options
  • Skills in data analysis, analytics and AI
  • Focus on user experience and accessibility
  • Transdisciplinary collaboration

“Vision: AI-powered dashboards will become the foundation of 21st-century public health surveillance, enabling rapid detection, prediction, and response to emerging health threats.”

Key References

  1. Parums, D.V. (2023). Editorial: Infectious Disease Surveillance Using Artificial Intelligence (AI) and its Role in Epidemic and Pandemic Preparedness. Medical Science Monitor, 29, e941209.

  2. Dong, E., Ratcliff, J., Goyea, T.D., et al. (2022). The Johns Hopkins University Center for Systems Science and Engineering COVID-19 Dashboard: data collection process, challenges faced, and lessons learned. The Lancet Infectious Diseases, 22(12), e370-e376.

  3. Ahn, E., Liu, N., Parekh, T., et al. (2021). A Mobile App and Dashboard for Early Detection of Infectious Disease Outbreaks: Development Study. JMIR Public Health and Surveillance, 7(3), e14837.

  4. Helminski, D., Kurlander, J.E., Renji, A.D., et al. (2022). Dashboards in Health Care Settings: Protocol for a Scoping Review. JMIR Research Protocols, 11(3), e34894.

Additional References

  1. Eddington, H.S., Trickey, A.W., Shah, V., Harris, A.H.S. (2022). Tutorial: implementing and visualizing machine learning (ML) clinical prediction models into web-accessible calculators using Shiny R. Annals of Translational Medicine, 10(24), 1414.

  2. Verma, H. (2024). Integrating Machine Learning Models in Shiny. Appsilon Blog. Retrieved from https://appsilon.com/integrating-machine-learning-models-in-shiny/

  3. Suvvari, T.K., Kandi, V. (2024). Artificial intelligence enhanced infectious disease surveillance - A call for global collaboration. New Microbes and New Infections, 62, 101494.

  4. World Health Organization. (2023). Global Influenza Surveillance and Response System (GISRS). Geneva: WHO Press.

Thank You

Questions

Contact Information:

  • Email: drkamarul@usm.my
  • Institution: School of Medical Sciences, Universiti Sains Malaysia