Resonant acoustic rheometry for assessing plasma coagulation in bleeding patients

Area of Science:

• Biomedical Engineering
• Hematology
• Rheology

Background:

• Life-threatening hemorrhage requires rapid hemostasis phenotyping.
• Current methods may lack the sensitivity needed for timely clinical decisions.
• Effective management of bleeding patients necessitates precise assessment of coagulation.

Purpose of the Study:

• To evaluate resonant acoustic rheometry (RAR) as a novel method for assessing plasma coagulation.
• To determine if RAR parameters can predict the need for blood product transfusion in bleeding patients.
• To explore the potential of machine learning models utilizing RAR data for clinical decision support.

Main Methods:

• Resonant acoustic rheometry (RAR) was used to measure plasma viscoelastic properties in 38 bleeding patients.
• Coagulation dynamics were assessed under varying activator and reagent conditions.
• A quadratic classifier was trained on RAR parameters to predict transfusion requirements.
• RAR results were correlated with traditional thromboelastography (TEG) parameters.

Main Results:

• RAR successfully captured dynamic plasma coagulation characteristics.
• RAR parameters showed correlations with TEG reaction time and functional fibrinogen, particularly when considering comorbidities.
• A machine learning model using RAR data achieved modest to high accuracy in predicting fresh frozen plasma and cryoprecipitate transfusions.
• The study demonstrated the feasibility of RAR for coagulation assessment.

Conclusions:

• Resonant acoustic rheometry is a feasible technique for assessing plasma coagulation in bleeding patients.
• RAR, combined with machine learning, shows potential for guiding transfusion decisions in critical care settings.
• Further validation with larger patient cohorts is necessary to establish RAR as a reliable, cost-effective hemostasis assay.