Evolve AI Institute

Extension Activity Project Guidelines & Templates

Lesson 11: AI in Healthcare – Diagnosis and Treatment

Table of Contents

  1. Healthcare AI Innovation Challenge
  2. Expert Interview Project
  3. Algorithmic Bias Investigation
  4. Healthcare AI News Monitor
  5. Cross-Curricular Integration Projects
  6. Assessment Rubrics for Extension Activities

Extension Activity 1: Healthcare AI Innovation Challenge

Duration: 2–3 weeks  |  Group Size: 3–4 students  |  Challenge: Design an AI solution to address a real healthcare problem

Learning Objectives

Students will: apply understanding of AI capabilities and limitations; practice design thinking; consider ethical implications; develop presentation skills; collaborate effectively in teams.

Project Guidelines

Phase 1: Problem Identification (Week 1)

Task: Identify a healthcare problem that could benefit from AI

Guiding Questions:

Requirements: Problem must be specific and clearly defined; should affect an identifiable group; students must research prior solutions; include data on scope and impact.

Deliverable: One-page problem statement including description, who is affected, current approaches and limitations, and why AI might help.

Phase 2: Solution Design (Week 2)

Task: Design a conceptual AI solution with attention to feasibility and ethics

Requirements cover five areas:

  1. AI Approach: Type of AI, data needed, training method, output/recommendations
  2. Implementation: Users, workflow, infrastructure, costs
  3. Expected Benefits: Patient impact, improvements, quantified benefits
  4. Ethical Considerations: Privacy, bias, accountability, equity, risks
  5. Safeguards: Testing, validation, human oversight, monitoring

Phase 3: Presentation and Feedback (Week 3)

Presentation Requirements: 5–7 minutes per group; visual aids; all team members participate; clear explanation of problem, solution, benefits, and ethics; respond to Q&A.

Structure: Hook (30 sec) → Problem with data (1 min) → AI solution (2 min) → Benefits (1 min) → Ethics and safeguards (1.5 min) → Conclusion (30 sec) → Q&A (3–5 min)

Project Templates

Template 1: Problem Statement

Team Members:

Problem Title:

1. Problem Description (What healthcare challenge are you addressing? 3–5 sentences)

2. Who Is Affected?

How many people are affected? (Provide data/estimates)

3. Current Approaches – How is this currently being addressed?

What are the limitations?

4. Why AI? (What AI capabilities are relevant?)

Has anyone tried AI for this before?

5. Significance – Why does this problem matter?

6. Research Sources (List 3–5 credible sources)

Template 2: Solution Design Worksheet

Team Members:

Project Title:

Part 1: AI Technical Approach

1. What type of AI will your solution use?

Machine Learning   Computer Vision   NLP   Predictive Analytics   Robotics   Other:

2. Describe how your AI system works:

3. What data does your AI system need?

Data Type 1: Source:

Data Type 2: Source:

Data Type 3: Source:

4. How would you train this AI system?

5. What does the AI output or recommend?

Part 2: Implementation

6. Who would use this system? Doctors Nurses Patients Researchers Administrators Other:

7. Describe a typical use scenario:

Step 1:

Step 2:

Step 3:

Step 4:

Step 5:

Part 3: Expected Benefits

10. Primary benefit to patients:

11. Additional benefits (2–3):

Part 4: Ethical Analysis

14. Privacy Concerns:

15. Bias Risks:

16. Accountability:

17. Equity and Access:

Part 5: Safeguards and Validation

19. Testing Plan:

20. Human Oversight: Can the AI be overridden by humans? Yes No

Part 6: Team Reflection

23. Biggest challenge in designing this solution:

24. Aspect you’re most proud of:

Template 3: Peer Feedback Form

Your Name:

Team Evaluated:

Project Title:

1. Problem Clarity (1–5): 1 2 3 4 5

Comments:

2. Solution Feasibility (1–5): 1 2 3 4 5

Comments:

3. Ethical Considerations (1–5): 1 2 3 4 5

Comments:

4. Potential Impact (1–5): 1 2 3 4 5

Comments:

5. Presentation Quality (1–5): 1 2 3 4 5

Comments:

6. What was the strongest aspect of this project?

7. What improvement should the team address?

8. What question do you still have about their solution?

9. Would you trust this AI system if it existed? Why or why not?

Extension Activity 2: Expert Interview Project

Duration: 2–3 weeks  |  Individual or Pair  |  Challenge: Interview a healthcare or technology professional about their AI experiences

Learning Objectives

Students will: practice professional communication and interview skills; gain authentic insights; synthesize information from primary sources; create multimedia content; connect classroom learning to real-world practice.

Project Phases

Phase 1 – Finding an Interview Subject: Potential subjects include doctors, nurses, medical technicians, medical school faculty, healthcare IT professionals, medical researchers, bioinformatics specialists, or public health officials. Use personal connections, school nurse contacts, local hospitals, universities, professional associations, or LinkedIn.

Phase 2 – Preparing for the Interview: Research the subject’s role and organization. Prepare 10–15 open-ended questions across five categories: Background and role (2–3), Current AI tools (3–4), Benefits and challenges (3–4), Ethical considerations (2–3), Future perspectives (2–3).

Phase 3 – Conducting the Interview: Arrive on time, dress appropriately, express gratitude, take detailed notes, ask follow-up questions, be respectful of time, ask permission before recording, send thank-you email within 24 hours.

Phase 4 – Creating Deliverable: Choose one format: Written Report (3–4 pages), Video Documentary (5–7 min), Podcast-Style Audio (8–10 min), or Multimedia Presentation (10–12 slides). All must include attribution, connection to 3+ lesson concepts, analysis (not just summary), and personal reflection.

Extension Activity 3: Algorithmic Bias Investigation

Duration: 2 weeks  |  Individual or Pair  |  Challenge: Research and analyze a documented case of bias in healthcare algorithms

Known Cases to Consider:

  1. Insurance algorithm underestimating illness severity in Black patients
  2. Diagnostic AI trained primarily on light-skinned patients
  3. Predictive models for kidney function including race as a variable
  4. Mental health AI showing different accuracy rates across groups
  5. Hospital allocation algorithms prioritizing certain populations

Project Phases:

  1. Case Selection & Research: Find 4–6 credible sources; understand technical details; identify affected populations; research how bias was discovered; investigate solutions.
  2. Root Cause Analysis: Examine training data, feature selection, historical patterns, development team, testing process, deployment context. Go beyond surface explanations.
  3. Impact Assessment: Document consequences at individual, community, institutional, and societal levels. Include both quantitative data and qualitative stories.
  4. Solutions Analysis: Evaluate technical fixes, policy changes, institutional reforms, and ongoing monitoring. Assess whether solutions address root causes.
  5. Presentation: Research Poster, Video Documentary (5–8 min), or Slide Presentation (12–15 slides).

Extension Activity 4: Healthcare AI News Monitor

Duration: Ongoing throughout semester  |  Weekly Rotation: 1–2 students per week  |  2–3 minute class presentation

Recommended Sources: STAT News, Healthcare IT News, medical journals (JAMA, NEJM, Lancet), MIT Technology Review, Science Daily, Nature News

Analysis Questions: What AI technology is involved? What problem does it address? Who developed it? What are benefits? What concerns exist? How does it connect to what we learned?

Presentation Structure: Headline/hook (10 sec) → Explain development (45 sec) → Why it matters (30 sec) → Class connections (30 sec) → Discussion question (15 sec)

Assessment: Pass/Fail completion credit. Excellence indicators include particularly timely selection, exceptional analysis, engaging discussion, and professional delivery.

Cross-Curricular Integration Projects

Project 5: Biology Connection – Genetic AI Analysis

Suitable for classes covering genetics, molecular biology, or human biology.

Students explore how AI analyzes genetic data to predict disease risk or guide treatment. Activities include researching ML identification of genetic variants, understanding polygenic risk scores, analyzing case studies of AI-enhanced genetic counseling, considering ethical issues in genetic prediction, and connecting to genetics concepts.

Project 6: Mathematics Connection – Diagnostic Accuracy Statistics

Suitable for statistics, probability, or data analysis units.

Students calculate and interpret measures of diagnostic accuracy. Activities include defining sensitivity/specificity/PPV/NPV, working with confusion matrices from published AI studies, calculating accuracy metrics, and exploring how prevalence affects predictive values (Bayesian reasoning).

Project 7: English/Language Arts – Healthcare AI Literature Analysis

Suitable for units on argumentation, research writing, or contemporary issues.

Students read and analyze articles about healthcare AI, then write argumentative essays. Possible prompts: Should AI diagnostics require FDA approval? Is it ethical for insurers to use AI predictions? Will healthcare AI help or hurt health equity? Should patients have the right to refuse AI involvement?

Assessment Rubrics for Extension Activities

Innovation Challenge Rubric (100 points)

Category (Points)Exemplary (A)Proficient (B)Developing (C)Beginning (D/F)
Problem Identification (20)Clear, specific, well-researched with dataAdequately defined with some researchBasic statement with limited researchVague or inappropriate; minimal research
AI Solution Design (25)Sophisticated understanding; detailed, technically soundSolid understanding; appropriate, feasibleBasic approach; some feasibility concernsWeak or inappropriate; not technically sound
Ethical Analysis (20)Comprehensive; thoughtful safeguards; deep understandingIdentifies major concerns; adequate safeguardsBasic consideration; limited safeguardsMinimal analysis; inadequate attention
Expected Impact (15)Compelling case; quantified benefits; realistic assessmentClear benefits; adequate justificationBasic benefits; limited justificationUnclear or unrealistic claims
Presentation (15)Highly professional; engaging; excellent visuals and coordinationProfessional; clear; good structureAdequate; some organizational issuesPoor; unclear; disorganized
Collaboration (5)Equal contribution; excellent teamworkGenerally equal; adequate teamworkUneven contribution or process issuesPoor teamwork; significant disparities

Expert Interview Rubric (100 points)

Category (Points)Exemplary (A)Proficient (B)Developing (C)Beginning (D/F)
Preparation (15)Excellent questions; thorough research; professionalGood questions; adequate researchBasic questions; limited researchPoor questions; inadequate preparation
Interview Quality (20)Insightful questions; active listening; rich responsesGood questions; adequate follow-upBasic questions; limited follow-upWeak questions; poor technique
Content Synthesis (25)Sophisticated synthesis; well-organized thematicallySolid synthesis; good organizationBasic synthesis; adequate organizationPoor synthesis; disorganized
Connection to Concepts (20)Insightful connections to multiple conceptsClear connections to 3+ conceptsBasic connections to 2–3 conceptsWeak or absent connections
Deliverable Quality (15)Exceptional; highly professional; engagingProfessional; meets all requirementsAdequate; meets basic requirementsPoor quality; missing elements
Reflection (5)Thoughtful, insightful reflectionAdequate reflectionBasic with limited depthMinimal or absent

Algorithmic Bias Investigation Rubric (100 points)

Category (Points)Exemplary (A)Proficient (B)Developing (C)Beginning (D/F)
Research Quality (20)Multiple high-quality sources; thorough; accurateAdequate sources; solid researchLimited sources; basic researchPoor sources; inadequate research
Case Understanding (15)Comprehensive; nuanced descriptionGood understanding; clear descriptionBasic understanding; adequate descriptionWeak; confused or inaccurate
Root Cause Analysis (25)Sophisticated; multiple causal factors; systemic understandingSolid analysis; identifies major causesBasic; identifies some causesSuperficial; misses key factors
Impact Assessment (15)Comprehensive; multiple levels; both quantitative and qualitativeGood analysis; adequate dataBasic analysis; limited dataWeak assessment; minimal evidence
Solutions Evaluation (15)Critical evaluation; considers adequacy; proposes improvementsAdequate evaluationBasic; limited critical analysisWeak or absent evaluation
Presentation (10)Highly effective; professional; engagingProfessional; clear; well-organizedAdequate; meets requirementsPoor; unclear or disorganized