Lesson 1Foundations of sepsis pathophysiology and clinical criteria (SIRS, qSOFA, SOFA, Sepsis-3)Reviews sepsis biology and host response, then links these mechanisms to bedside signs such as hypotension, tachycardia, and organ dysfunction. Compares SIRS, qSOFA, SOFA, and Sepsis-3 criteria and their use in ED triage.
Host response to infection and organ dysfunctionHemodynamic changes and microcirculatory failureSIRS, qSOFA, SOFA: components and thresholdsSepsis-3 definition and septic shock criteriaLimitations of clinical scores in the EDLesson 2Safety, failure modes, and mitigation: false positives/negatives, model drift, data quality issues, adversarial inputsIdentifies safety risks such as false positives, false negatives, model drift, and poor data quality. Explores adversarial or unexpected inputs, robust monitoring, guardrails, human oversight, and processes for safe model updates.
False positives, false negatives, and harm modesData quality checks and anomaly detectionModel drift, recalibration, and retrainingAdversarial or unexpected input handlingHuman oversight, overrides, and governanceLesson 3Evaluation metrics and validation strategies for sepsis prediction: AUROC, AUPRC, calibration, lead time, decision curve analysisDefines key performance metrics for sepsis prediction, including AUROC, AUPRC, calibration, and lead time. Explains internal and external validation, temporal validation, and decision curve analysis for clinical utility assessment.
AUROC, AUPRC, and class imbalanceCalibration curves and risk stratificationLead time and horizon-specific performanceInternal, external, and temporal validationDecision curve analysis and net benefitLesson 4Feature engineering and temporal modelling: time-series preprocessing, sliding windows, trend extractionExplains how to clean and align ED time-series data for modelling. Covers resampling, handling irregular intervals, sliding windows, trend and variability features, and encoding interventions and clinical context over time.
Time alignment, resampling, and interpolationSliding windows and prediction horizonsTrend, variability, and derivative featuresEncoding interventions and care escalationHandling irregular and sparse time-seriesLesson 5Machine learning models for acute risk prediction: logistic regression, gradient boosted trees, RNNs, temporal convolutional networks, transformer-based time-seriesCompares modelling approaches for acute sepsis risk prediction, from logistic regression to gradient boosted trees and deep sequence models. Highlights strengths, limitations, interpretability, and suitability for ED time constraints.
Logistic regression and regularization choicesGradient boosted trees and feature importanceRecurrent neural networks for sequencesTemporal convolutional networks for time-seriesTransformers for clinical time-series dataLesson 6Data modalities for real-time sepsis detection: vitals, labs, nursing notes, medication, waveformsDescribes key real-time data streams in the ED, including vital signs, laboratory tests, medications, nursing documentation, and physiologic waveforms. Discusses sampling rates, reliability, and how each modality signals evolving sepsis.
Vital signs and continuous monitoring feedsLaboratory panels, cultures, and turnaround timesMedication orders, fluids, and vasopressorsNursing notes, triage text, and flowsheetsWaveforms from monitors and bedside devicesLesson 7Integration with ED workflows and EHR systems: event streams, FHIR, HL7, SMART on FHIR apps, CDS HooksDescribes how AI sepsis models integrate into ED workflows and EHRs. Reviews event streams, HL7, FHIR resources, SMART on FHIR apps, and CDS Hooks, emphasising usability, reliability, and minimal disruption to clinical practice.
Event-driven architectures and data streamsHL7 and FHIR resources for sepsis signalsSMART on FHIR apps for bedside decision supportCDS Hooks for context-aware recommendationsWorkflow mapping and usability testingLesson 8Clinical alert design and human factors: thresholds, alarm fatigue mitigation, escalating workflows, who receives alertsCovers principles of alert design for ED clinicians, including threshold selection, tiered alerts, and routing to appropriate roles. Addresses alarm fatigue, alert timing, escalation pathways, and presenting explanations and context.
Choosing thresholds and alert tiersAlarm fatigue and suppression strategiesWho receives alerts and on which channelsEscalation workflows and handoff supportExplaining alerts and providing contextLesson 9Regulatory and evidence requirements for diagnostic AI: FDA/CMS considerations, clinical validation study design, prospective pilots, reporting standards (TRIPOD, CONSORT-AI)Outlines regulatory and evidence expectations for diagnostic AI in sepsis, including FDA pathways, CMS considerations, and clinical validation. Reviews prospective pilots and reporting standards such as TRIPOD and CONSORT-AI.
FDA pathways for diagnostic support toolsCMS, reimbursement, and quality programsDesigning robust clinical validation studiesProspective pilots and phased rolloutsTRIPOD and CONSORT-AI reporting guidanceLesson 10Deployment frequency and latency considerations: near-real-time streaming vs batch scoring, handling missing and delayed dataDiscusses deployment architectures for sepsis models, comparing near-real-time streaming with batch scoring. Addresses latency budgets, handling missing or delayed data, backfilling, and monitoring data pipeline health in the ED.
Near-real-time streaming vs batch scoringLatency budgets and SLA definitionsImputation for missing and delayed inputsBackfilling, replay, and late-arriving dataMonitoring pipelines and system resilience
