Related Experiment Video
Updated: Sep 6, 2025
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
Published on: April 13, 2013
Exemplar deep and hand-modeled features based automated and accurate cerebral hemorrhage classification method
M Sait Din1, Sukru Gurbuz1, Erhan Akbal2
1Department of Emergency, College of Medicine, Inonu University, Malatya, Turkey.
Insights
A machine learning model accurately classifies cerebral hemorrhage (CH) types from CT scans using fused features. This approach enhances diagnostic speed and accuracy for emergency medical services.
Area of Science:
- Medical Imaging
- Machine Learning
- Neurology
Background:
- Cerebral hemorrhage (CH) diagnosis is critical for effective treatment.
- Prompt and accurate CH classification is essential in clinical settings.
Purpose of the Study:
- To develop an accurate machine learning model for classifying cerebral hemorrhage (CH) types.
- To simplify the diagnostic process for CH.
Main Methods:
- Utilized a dataset of 9818 CT images across five categories.
- Developed an exemplar fused feature generator combining AlexNet, LBP, and LPQ.
- Employed Neighborhood Component Analysis (NCA) for feature selection and Support Vector Machine for classification.
Main Results:
- The model achieved high accuracies, including 97.47% with five hold-out validations and 96.34% with ten-fold cross-validation.
- Demonstrated robust performance across multiple validation methods.
Conclusions:
- The developed model and feature generation technique are successful and robust.
- This model can significantly aid emergency services in prompt CH diagnosis.
Background:
Cerebral hemorrhage (CH) is a commonly seen disease, and an accurate diagnosis of the type of CH is a very crucial step in treatment. Therefore, CH requires a prompt and accurate diagnosis. To simplify this process, an accurate CH classification model is presented using a machine learning technique.
Material And Method:
A computed tomography (CT) image dataset was collected retrospectively in this research. This dataset contains 9818 images with five categories. An exemplar fused feature generator is presented to classify these features. This generator uses pre-trained AlexNet, local binary pattern (LBP), and local phase quantization (LPQ). The neighborhood component analysis (NCA) method selects the top features, and the chosen feature vector is classified on the support vector machine.
Results:
Six validation methods are utilized to calculate the performance of the presented exemplar fused features and NCA-based CH classification model. This model attained 97.47%, 96.05%, 95.21%, 93.62%, 91.28% and 96.34% accuracies using five hold-out validations and ten-fold cross-validation respectively.
Conclusions:
The calculated results clearly demonstrate the success and robustness of the introduced exemplar fused feature generation and NCA-based model. Furthermore, this model can be used in emergency services to overcome a prompt diagnosis of CH.
