Related Experiment Video
Updated: Aug 6, 2026
14:08
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
Published on: April 13, 2013
Development and Internal Validation of an Automated CT-Based Model for Early Mortality Prediction in Traumatic Brain
Sujong Shin1, Jaewoo Chung2, Jungchan Cho3
1Department of AI-based Convergence, Dankook University, Yongin, Gyeonggi-do, Republic of Korea.
World Neurosurgery
|July 23, 2026
Summary
An automated framework using deep learning to analyze CT scans for traumatic brain injury (TBI) shows promise. This tool accurately predicts early mortality risk, comparable to traditional methods, especially when combined with clinical data.
Area of Science:
- Neuroscience
- Radiology
- Medical Imaging
Background:
- Traumatic brain injury (TBI) outcomes are significantly affected by intracranial hemorrhage, midline shift, and hypoxic injury.
- Current prognostic tools like the Marshall CT score, Rotterdam CT score, and IMPACT model require manual interpretation, limiting consistency and accessibility.
Purpose of the Study:
- To develop a fully automated multimodal framework for early mortality prediction in TBI patients.
- To compare the prognostic performance of this automated framework with conventional TBI scoring systems.
Main Methods:
- Deep learning models were used to automatically extract hematoma volume and midline shift (MLS) from CT images.
- A shallow multilayer perceptron (MLP) predictor integrated these imaging biomarkers with clinical variables for mortality prediction.
- SHAP analysis was employed for model interpretability to understand feature contributions.
Main Results:
- Automated imaging biomarkers demonstrated discriminative performance comparable to conventional TBI scoring systems.
- Integrating automated imaging biomarkers with clinical variables further improved prediction performance.
- The developed framework achieved similar prognostic accuracy to established scoring systems.
Conclusions:
- Automated hematoma volume and MLS provide prognostic performance on par with conventional TBI scoring systems.
- Combining automated imaging biomarkers with clinical variables enhances early mortality risk stratification in TBI.
- A fully automated and reproducible framework for TBI risk stratification is feasible.
