AI Patient Simulator (SOOS Oral Exam Practice)

What is SOOS?

SOOS (Structured Oral Objective Examination) is a type of oral practical exam used in medical education where residents interact with standardized patients (actors) who simulate specific clinical scenarios. The standardized patient also acts as the evaluator, scoring the resident's:

• Clinical history-taking skills
• Physical examination technique
• Diagnostic reasoning
• Communication and professionalism
• Patient education and management planning

The Challenge with Traditional SOOS Practice

Limited practice opportunities:

• Standardized patients are expensive to hire ($50-100/hour)
• Requires scheduling and coordination
• Only available during formal teaching sessions
• Limited scenarios (budget constraints)

Inconsistent evaluation:

• Different standardized patients interpret rubrics differently
• Subjective scoring can vary
• Limited immediate feedback

Scheduling difficulties:

• Residents work unpredictable hours
• Hard to find practice partners
• Formal OSCE practice sessions are rare (2-3 times/year)

Anxiety and stress:

• High-stakes with limited preparation
• Residents feel underprepared
• No safe space to make mistakes and learn

Noumaris Learn Solution: AI Patient Simulator

An AI-powered voice conversation system that plays the role of a patient in a medical encounter. Practice unlimited scenarios, get instant evaluation, and build clinical confidence.

How It Works

Step 1: Choose a Scenario

Available SOOS Scenarios:
┌─────────────────────────────────────────┐
│ 🩺 Chest Pain (Difficulty: Medium)     │
│    74-year-old male, acute chest pain  │
│    Focus: History-taking, cardiac exam │
│    Time: 10 minutes                     │
│    [Start Practice]                     │
└─────────────────────────────────────────┘

┌─────────────────────────────────────────┐
│ 🤰 Pregnant Patient - Abdominal Pain   │
│    28-year-old, 24 weeks pregnant      │
│    Focus: OB history, differential dx  │
│    Time: 12 minutes                     │
│    [Start Practice]                     │
└─────────────────────────────────────────┘

Step 2: Have a Natural Conversation

AI Patient (via voice):
"Hello doctor. Thanks for seeing me today. I've been having
this terrible chest pain that started about 2 hours ago."

Resident (speaks into phone/computer):
"I'm sorry to hear that. Can you describe the pain for me?
Where exactly do you feel it?"

AI Patient:
"It's right here in the middle of my chest [touches chest].
It feels like someone is squeezing really hard."

Resident:
"Does the pain go anywhere else? Down your arm or to your jaw?"

AI Patient:
"Yes! Now that you mention it, my left arm feels kind of tingly."

Step 3: Get Real-Time Evaluation After the encounter, receive detailed feedback:

SOOS Evaluation Report
Scenario: Acute Chest Pain (Cardiac)
Score: 82/100

✅ STRENGTHS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
✓ Excellent opening (introduced self, explained purpose)
✓ Systematic pain history (OPQRST covered)
✓ Identified key red flags (crushing pain, left arm radiation)
✓ Appropriate urgency recognized

⚠️  AREAS FOR IMPROVEMENT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚠ Missed asking about cardiac risk factors (diabetes, smoking)
⚠ Did not screen for other symptoms (nausea, diaphoresis)
⚠ Could improve empathy statements (patient seemed anxious)

📊 RUBRIC BREAKDOWN
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
History-taking:        18/20 ⭐⭐⭐⭐
Physical Exam:         14/20 ⭐⭐⭐
Diagnosis:             16/20 ⭐⭐⭐⭐
Communication:         14/20 ⭐⭐⭐
Management:            20/20 ⭐⭐⭐⭐⭐

💡 LEARNING POINTS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
• For cardiac chest pain, always ask: "OLDCARTS + Risk Factors"
• Patient appeared anxious - could use phrases like "I understand
  this is concerning" to build rapport
• Consider asking "What worries you most?" to address patient
  concerns

🎯 NEXT STEPS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Practice recommended scenarios:
1. "MI vs Angina" - Sharpen diagnostic differentiation
2. "Anxious Patient - Chest Pain" - Build communication skills

Key Features

1. Natural Voice Interaction

• Speak naturally: No typing, just talk like a real patient encounter
• Realistic responses: AI patient responds in character (age, personality, anxiety level)
• Interruption handling: Can interrupt or ask clarifying questions
• Accent/language options: Practice with diverse patient populations

2. Comprehensive Scenario Library

Categories:

• Acute Care: Chest pain, abdominal pain, shortness of breath, altered mental status
• Chronic Disease: Diabetes management, hypertension follow-up, COPD exacerbation
• Preventive Care: Well-child visits, cancer screening counseling, immunizations
• Mental Health: Depression screening, anxiety, substance use counseling
• Women's Health: Contraception counseling, abnormal bleeding, prenatal care
• Pediatrics: Fever in infant, asthma, developmental concerns

Difficulty Levels:

• Beginner: Straightforward presentations, cooperative patients
• Intermediate: Complex medical histories, requires systematic approach
• Advanced: Challenging patients (angry, anxious, poor historians), diagnostic uncertainty

3. Official Exam Rubrics

Evaluation based on actual SOOS/OSCE rubrics from:

• Canadian specialty colleges (CFPC, Royal College)
• Medical school OSCEs (U of T, UBC, McGill)
• US Step 2 CS equivalent evaluations

4. Adaptive Difficulty

• Learns your level: Suggests appropriate scenarios based on performance
• Progressive challenges: As you improve, unlocks harder scenarios
• Weak area focus: Identifies communication gaps, history-taking errors

5. Performance Tracking

• Scenario completion history: See all past encounters and scores
• Skill progression: Track improvement over time
• Rubric mastery: Visual dashboard of strengths/weaknesses
• Readiness assessment: "You're 85% ready for OSCE based on performance"

Use Cases

Use Case 1: OSCE Preparation (Medical Student)

Emma, Year 3 Medical Student, OSCE exam in 4 weeks

Challenge:

• Never done an OSCE before, anxious about format
• Only 2 practice sessions scheduled before exam
• Wants to practice but classmates are busy

Solution:

1. Starts with "Beginner" scenarios to build confidence
2. Practices 3-4 scenarios per week (30 min each)
3. Reviews AI feedback, focuses on weak areas (patient rapport)
4. Repeats challenging scenarios until comfortable
5. Week before exam: runs through all exam stations for timing practice

Outcome:

• Feels confident and prepared for OSCE
• Scores 88/100 on actual exam (class average 76)
• Develops systematic approach to patient encounters

Use Case 2: Rusty Skills Refresher (Attending Physician)

Dr. Kim, Family Physician, 10 years post-residency

Challenge:

• Mostly doing admin work now, fewer direct patient encounters
• Needs to do recertification OSCE
• Feels rusty on physical exam skills and communication

Solution:

1. Uses AI patient simulator to practice common scenarios
2. Focuses on communication skills (empathy, shared decision-making)
3. Gets feedback on missed examination steps
4. Practices 2-3 scenarios per week for 8 weeks

Outcome:

• Passes recertification OSCE comfortably
• Feels more confident in patient interactions
• Identifies continuing education needs (dermatology weak area)

Use Case 3: International Medical Graduate (IMG)

Dr. Patel, IMG preparing for Canadian licensing exams

Challenge:

• Trained in different healthcare system (India)
• Canadian communication style different (more patient-centered)
• Needs to pass NAC OSCE for licensure
• Limited access to standardized patients for practice

Solution:

1. Practices "Canadian-style" patient interactions
2. Learns cultural differences (e.g., addressing substance use, end-of-life)
3. AI provides feedback on communication style
4. Unlimited practice without scheduling hassles
5. Focuses on "difficult conversations" scenarios (bad news, refusal of treatment)

Outcome:

• Passes NAC OSCE on second attempt (failed first time)
• Feels more comfortable with Canadian patient communication
• Identifies specific cultural learning needs

Use Case 4: Residency Program - Structured Practice

McMaster Family Medicine Program (80 residents)

Challenge:

• Residents have variable OSCE performance
• Limited standardized patient budget ($10,000/year)
• Program director wants consistent evaluation practice

Solution:

1. Program licenses AI patient simulator for all residents
2. Assigns mandatory weekly practice scenarios
3. Tracks resident progress via admin dashboard
4. Identifies residents who need extra support
5. Uses in-person standardized patients only for high-stakes formative OSCEs

Outcome:

• OSCE pass rate improves 82% → 94%
• Reduces standardized patient costs by 60%
• Residents report feeling better prepared
• Earlier identification of struggling learners

AI Patient Personalities & Scenarios

AI patients have realistic personalities and backgrounds:

Example 1: Anxious Patient

Name: Susan Chen
Age: 42
Presentation: "I'm really worried about this lump in my breast..."
Personality: Anxious, tearful, needs reassurance
Teaching point: Empathy, addressing fears, shared decision-making

Example 2: Angry Patient

Name: Robert Johnson
Age: 68
Presentation: "I've been waiting 6 months for this specialist appointment!"
Personality: Frustrated, irritable, interruptive
Teaching point: De-escalation, active listening, validation

Example 3: Poor Historian

Name: Maria Gonzalez
Age: 74 (via interpreter)
Presentation: Vague abdominal pain, language barrier
Personality: Quiet, defers to family, cultural considerations
Teaching point: Working with interpreters, systematic questioning

Example 4: Adolescent Patient

Name: Tyler Williams
Age: 16
Presentation: "My mom made me come... I'm fine."
Personality: Reluctant, minimizes symptoms, protective of privacy
Teaching point: Building rapport with teens, confidentiality discussion

Technical Innovation (Simplified Explanation)

Speech-to-Text:

• Converts your voice to text in real-time
• Understands medical terminology
• Handles accents and speech patterns

AI Conversation Engine:

• Understands context and intent (not just keywords)
• Stays in character throughout encounter
• Responds naturally, like a real patient would

Text-to-Speech:

• Converts AI response to natural-sounding voice
• Different voices for different patient demographics
• Emotional tone (anxious, calm, angry)

Evaluation Engine:

• Compares your conversation to rubric criteria
• Identifies missed questions, communication gaps
• Generates personalized feedback and learning points

Integration with Flashcard System

Synergy between tools:

1. Fail a SOOS scenario on diabetes management
2. System generates targeted flashcard deck on diabetes
3. Master the flashcards over next week
4. Retry the scenario with improved knowledge
5. Pass with higher score, confidence improves

Example workflow:

SOOS Practice: "Type 2 Diabetes Follow-Up" → Score 65/100
    ↓
AI Analysis: "Weak on medication management (metformin, GLP-1)"
    ↓
Auto-generates: "Diabetes Pharmacology" deck (15 cards)
    ↓
Study flashcards for 3 days
    ↓
Retry scenario → Score 88/100
    ↓
✓ Mastered! Move to next topic.

Pricing

Included in Noumaris Learn subscription:

• Individual: $29-49/month (includes flashcards + simulator)
• Institutional: $15-25/resident/month

Usage limits:

• Individual: Unlimited scenarios
• Institutional: Unlimited scenarios + admin dashboard

Roadmap

Phase 1 (Launch)

• 20 core scenarios (cardiology, respiratory, GI, mental health)
• Voice conversation in English
• Standard rubric evaluation
• Basic performance tracking

Phase 2 (Months 3-6)

• 100+ scenarios across all specialties
• Multi-language support (French, Mandarin, Spanish)
• Video integration (patient facial expressions, body language)
• Peer comparison (how you rank vs other learners)

Phase 3 (Months 6-12)

• Custom scenario creation (upload your own case)
• Virtual physical exam (VR integration)
• Multi-patient scenarios (prioritization, triage)
• Team-based assessments (work with AI team members)

Phase 4 (Year 2+)

• VR headset support (full immersive patient encounter)
• Haptic feedback (palpation, percussion simulation)
• Objective structured clinical examination (full OSCE station with multiple components)

Success Metrics

For Learners:

• OSCE pass rate: 85%+ (vs 70% baseline)
• Confidence scores: 4.5/5 (self-reported readiness)
• Practice volume: 12+ scenarios before real exam (vs 2-3 with traditional)

For Programs:

• Standardized patient cost reduction: 50-70%
• Earlier identification of struggling learners
• More consistent evaluation practice

Comparison to Alternatives

Solution	Cost	Availability	Realism	Evaluation Quality
Standardized Patients	$50-100/hr	Limited (scheduled)	⭐⭐⭐⭐⭐ High	⭐⭐⭐⭐ Variable
Peer Practice	Free	Hard to schedule	⭐⭐ Low	⭐ Inconsistent
Online Videos	Free-$50	Unlimited	⭐⭐ Passive	⭐ None
VR Simulators	$5,000+	Limited (institutions)	⭐⭐⭐⭐ High	⭐⭐⭐ Good
Noumaris AI Simulator	$29-49/mo	Unlimited	⭐⭐⭐⭐ Very Good	⭐⭐⭐⭐ Excellent

Privacy & Ethics

Conversations are confidential:

• Practice sessions are not shared without permission
• Institutions can see aggregated data, not individual transcripts
• Option to practice "anonymously" (no tracking)

AI limitations disclosed:

• Clear labeling: "AI simulation for practice, not real patient care"
• Disclaimer: "Supplement to, not replacement for, standardized patients"
• Human review: Evaluation rubrics validated by medical educators

Cultural sensitivity:

• Diverse patient scenarios (age, gender, ethnicity, cultural backgrounds)
• Disability representation (patients with hearing loss, mobility issues)
• LGBTQ+ inclusive scenarios

---

See also:

• Learn Platform Overview
• Intelligent Flashcard System