CT ProVision
Predicting severe outcome of Subarachnoid Hemorrhage patients parting from an emergency room CT. This web app calculates probability of intra-hospital death and vasospasm based on CT, Hunt-Hess and WFNS data.
This web page is a designed as a Supporting Material to a publication submitted to Neurology, manuscript number NEUROLOGY/2018/910448. The method is patent pending WO2019025656 and PCT/ES2018/070539. The page while being in beta version, is under constant developement. If you get the error, contact developer providing submitted clinical data and patient ID submitted.
The method is filed with patent application in national phases in European Patent Office and in United States of America.
Login / Register »
About the biomarker panel model
The biomarker panel model is a method that allows to predict the likelihood of itra-hospital death or vasospasm in patients suffering from subarachnoid hemorrhage.
Learn more »
What the model can achieve better than individual biomarkers?
Currently Hunt-Hess and WFNS are used as biomarkers.
Hunt-Hess has ROC/AUC of 0.725 for intra-hospital death and 0.625 for vasospasm.
WFNS has ROC/AUC of 0.763 for intra-hospital death and 0.641 for vasospasm.
Our biomarker model developed on 85 patients has ROC/AUC of 0.907 for intra-hospital death and 0.653 for vasospasm. That translates for intra-hospital death to 75% sensitivity at 90% specificity for the model, compared to 31% for WFNS and 25% for Hunt-Hess at 90% specificity.
Learn more »
Try it and join us.
Test your SAH CT images with our web app. For MDs who will supply ten or more images we agree to provide authorship in validation paper.
Learn more »