Detection of Subarachnoid Hemorrhage in Computed Tomography Using Association Rules Mining

Area of Science:

• Neurology
• Medical Imaging
• Data Science

Background:

• Subarachnoid hemorrhage (SAH) is a critical type of stroke, often caused by ruptured intracranial aneurysms.
• Computed tomography (CT) scans are standard for SAH diagnosis but can be challenging for identifying abnormalities.
• Early detection of SAH is crucial for timely therapeutic intervention.

Purpose of the Study:

• To develop an automated system for detecting subarachnoid hemorrhage (SAH) from CT images.
• To enhance the accuracy and efficiency of SAH diagnosis.
• To compare a novel NCFP-growth algorithm with existing methods.

Main Methods:

• Feature extraction from CT images using the gray-level cooccurrence matrix (GLCM).
• Application of the New Association Classification Frequent Pattern (NCFP-growth) algorithm based on association rules.
• Comparative analysis with FP-growth algorithms (with and without association rules).

Main Results:

• The proposed NCFP-growth algorithm demonstrated superior classification accuracy.
• The automated detection system achieved an accuracy rate of 95.2%.
• The NCFP-growth approach outperformed conventional data mining algorithms.

Conclusions:

• The developed automated detection method shows significant promise for improving SAH diagnosis.
• The NCFP-growth algorithm offers a more accurate approach for identifying SAH in CT scans.
• This technique can aid in earlier recognition and treatment of subarachnoid hemorrhage.