Intelligent computational ensemble model for predicting cerebral aneurysm using the concept of region localization in multi-section CT angiography

Intracranial aneurysms and hemorrhage, also known as cerebral or brain aneurysms, are abnormal bulges or ballooning in the walls of blood vessels in the brain. When intracranial aneurysm leaks or ruptures, it results in bleeding within the brain, referred to as a hemorrhagic stroke. Multi-section computed tomography angiography (CTA) data is used to study because it is an effective technique for identifying cerebral atherosclerosis, arterial and intracranial hemorrhage. The hybrid model RID_Net is proposed that integrates ResNet15, InceptionV3, and DenseNet201 architectures in collectively, optimizing for precision through individual efficacy testing of each model. This ensemble model enhances feature extraction and classification performance, demonstrating the effectiveness of combining multiple transfer learning paradigms in biomedical image analysis. The performance is validated by comparing the proposed model with existing studies based on metrics such as accuracy (%), precision (%), recall (%) and f1-score (%).