Machine Learning and Intracranial Aneurysms: From Detection to Outcome Prediction

Area of Science:

• Biomedical Sciences
• Medical Imaging
• Artificial Intelligence

Background:

• Machine learning (ML) offers advanced capabilities for data analysis in biomedical research.
• Intracranial aneurysms (IA) present complex diagnostic and prognostic challenges.
• Traditional statistical methods have limitations in handling large, complex medical datasets.

Purpose of the Study:

• To review the application of ML in intracranial aneurysm (IA) research.
• To assess the performance of ML models in IA detection, rupture prediction, and outcome assessment.
• To evaluate the potential of ML in improving IA patient management.

Main Methods:

• Review of studies applying ML algorithms, including convolutional neural networks, to IA data.
• Analysis of ML model performance in tasks such as segmentation, classification, and prediction.
• Comparison of ML approaches with traditional statistical methods and expert analysis.

Main Results:

• ML models, particularly convolutional neural networks, demonstrate high sensitivity in IA detection, often exceeding expert performance.
• ML shows potential in differentiating ruptured from unruptured IAs and predicting rupture risk.
• ML-based prediction of complications and functional outcomes post-subarachnoid hemorrhage (SAH) is more established than for unruptured IA procedures.

Conclusions:

• Machine learning holds significant potential to revolutionize intracranial aneurysm patient management.
• Current clinical translation of ML in IA research is limited, indicating a need for further development and validation.
• ML algorithms offer superior performance over conventional methods for various IA-related predictive tasks.