Types of AI in Healthcare

1. Predictive/Deterministic AI

Often rule-based or uses traditional machine learning to predict or detect specific outcomes.
Use cases:
- Medication dosing (e.g., insulin calculators)
- Clinical protocols (e.g., Wells Criteria for PE)
- Vital sign alerts (e.g., sepsis alarms based on thresholds)
Examples:
- Wells Criteria (MDCalc) – rule-based scoring
- NEJM study – probably refers to predictive models used in clinical trials or outcome prediction
- Epilepsy detection using EEG analysis – pattern recognition

Generates new content based on learned patterns.
Use cases:
- Documentation assistance (e.g., clinical notes from conversations)
- Clinical summaries (e.g., summarizing EHRs or patient histories)
Examples:
- AI scribe tools like Abridge – listens to clinician-patient convo and drafts notes
- Chatbots for history intake or literature summarization

SOAP note generator for patient encounters - Using GenAI
AI-powered EHR generator or EHR Summarizer
Patient Intake ChatBot - Automate patient history intake via chat or voice, then summarize for the clinician
Healthcare x Finance (Study on healthcare drug industrial leaders like Elli Lilly, Novo Nordisk’s weight loss, diabetes drugs and how if affects their market sentiment)
Emergency Room (ER) Triage Assistant - Prioritize ER patients based on symptom severity using AI