Mortality-SAH is an AI-powered clinical tool designed to estimate the risk of mortality within the first 90 days following aneurysmal subarachnoid hemorrhage (aSAH). Built on state-of-the-art deep learning techniques, it relies solely on non-contrast computed tomography (CT) scans acquired at the time of admission—providing fast, fully automated prognostic insight when decisions are most critical.
This tool is the result of a collaborative project between GEIBAC and the University of Augsburg (Germany), combining clinical expertise with cutting-edge artificial intelligence research.
It is currently undergoing multicenter validation across several Spanish hospitals.
Key Features
Automated Data Processing
Converts DICOM images into NIfTI format for streamlined analysis.
User-Friendly Input Structure
Simply organize your DICOM images into patient-specific folders, and the system will handle preprocessing and inference.
Output Files
Processed NIfTI images & Prediction report (predictions.csv) summarizing risk probabilities
How it Works
Unlike traditional prognostic scores that require manual input and clinical variables, Mortality-SAH leverages convolutional neural networks trained on large datasets of de-identified CT scans. It provides a mortality probability score directly from imaging, eliminating user bias and reducing evaluation time.
Publications
-
2023 Brain Sciences
García-García S. et al.
Mortality Prediction of Patients with Subarachnoid Hemorrhage Using a Deep Learning Model Based on an Initial Brain CT Scan
-
2023 Brain and Spine
García-García S. et al.
Neurological Outcome Prediction in Patients with Subarachnoid Hemorrhage Using a Model Based on Initial CT Scan, Clinical Data and Neural Networks
-
2024 World Neurosurgery
García-García S. et al.
A Fully Automated Pipeline Using Swin Transformers for Deep Learning-Based Blood Segmentation on Head Computed Tomography Scans After Aneurysmal Subarachnoid Hemorrhage
-
2023 Neurocirugía
García-García S. et al.
Mortality prediction in patients with subarachnoid hemorrhage using an artificial intelligence model based on initial CT scan and neural networks
License
This software is distributed under the Creative Commons Attribution- NonCommercial (CC BY-NC) license. It is free to use, modify, and share for educational, personal, or non-profit purposes. Commercial use is strictly prohibited without prior written permission.