Lesson 1Foundations of sepsis pathophysiology and clinical criteria (SIRS, qSOFA, SOFA, Sepsis-3)Looks at sepsis biology and how the body responds, linking these to signs doctors see like low blood pressure, fast heart rate, and organ problems. Compares SIRS, qSOFA, SOFA, and Sepsis-3 criteria and how they help sort patients in emergency 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 inputsPoints out safety risks like wrong positives or negatives, model changes over time, bad data quality. Looks at tricky or surprise inputs, strong monitoring, safety barriers, doctor oversight, and safe ways to update models.
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 analysisExplains main performance measures for sepsis prediction like AUROC, AUPRC, how well calibrated, and warning time ahead. Covers checking inside and outside data, time-based checks, and decision curves to see real clinical value.
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 modeling: time-series preprocessing, sliding windows, trend extractionShows how to clean and line up emergency department time-series data for models. Includes re-sampling, dealing with uneven times, sliding windows, pulling out trends and changes, and noting treatments and clinical background 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 ways to model sudden sepsis risk, from simple logistic regression to boosted trees and deep sequence models. Notes strengths, weaknesses, how easy to understand, and fit for quick emergency times.
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 main real-time data flows in emergency departments, like vital signs, lab tests, meds, nurse notes, and body signal waves. Talks about how often sampled, reliability, and how each shows sepsis developing.
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 HooksExplains how AI sepsis models fit into emergency workflows and patient record systems. Reviews event flows, HL7, FHIR parts, SMART on FHIR apps, and CDS Hooks, stressing ease of use, reliability, and little change to daily 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 how to design alerts for emergency doctors, picking thresholds, stepped alerts, and sending to right people. Deals with alert tiredness, timing, step-up paths, and giving reasons and background.
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 what regulators and evidence need for diagnostic AI in sepsis, like FDA paths, CMS views, and clinical checks. Reviews forward pilots and reporting like 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 ways to deploy sepsis models, near-real-time streams vs batch checks. Covers time limits, missing or late data, filling gaps, and watching data flows in emergency departments.
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
