PROJECT TEAM
Benjamin Skov Kaas-Hansen, MD, MSc, PhD, Rigshospitalet
Hans-Christian Thorsen-Meyer MD, PhD, Rigshospitalet
Davide Placido MSc, Rigshospitalet
Anders Perner Professor, MD, PhD, Rigshospitalet
THE NEED
Each year, around 30,000 patients are admitted to ICUs in Denmark, and half require mechanical ventilation. Accurately predicting which patients need prolonged MV is critical, but current decisions rely on clinician judgment, which often fails due to bias and limited data. No reliable, data-driven tool exists to guide this key intervention.
THE SOLUTION
trachAI leverages deep learning to predict prolonged MV needs using data from a wide range of ICU parameters. The system is accessible via a simple web-based interface requiring only minimal data input, making it fast and practical for daily clinical use. It continuously learns and calibrates based on actual outcomes, ensuring accuracy improves over time. The tool can also be integrated directly into electronic patient records for seamless use.
trachAI: AI-informed identification of tracheostomy need - Rigshospitalet
Call 3 - 2022
500.000 DKK
Clinical Area
Intensive care, Pulmonology
Technology
AI Decision Support Tool
PROJECT SUMMARY
trachAI is an AI-powered clinical decision support tool that helps intensive care physicians identify patients who are likely to need prolonged mechanical ventilation (MV) and would therefore benefit from a tracheostomy.
CLINICAL IMPACT
trachAI enables more precise use of tracheostomy, a procedure that allows patients to be more comfortable, awake, and mobile during mechanical ventilation. With better predictions, patients can avoid unnecessary risks from delayed or unnecessary interventions. This supports better recovery, shorter ICU stays, and more efficient use of healthcare resources—ultimately leading to improved patient outcomes and quality of care.