Computer-Assisted Adjuncts for Aneurysmal Morphologic Assessment: Toward More Precise and Accurate Approaches.

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Biology

Background:

• Intracranial aneurysms (IAs) pose significant risks, with treatment decisions heavily reliant on morphological measurements from 2D angiographic images.
• Manual 2D measurements are prone to inaccuracies due to the complex 3D geometry of IAs and inter-observer variability.
• These inaccuracies can lead to suboptimal treatment strategies and potentially adverse patient outcomes.

Purpose of the Study:

• To develop and evaluate AView, an image-based computational tool designed to enhance the accuracy and consistency of IA morphological analysis.
• To quantify the precision of AView's computer-assisted measurements against known standards and compare them with traditional manual methods.
• To assess the clinical utility of AView in improving measurement consistency among clinicians for intracranial aneurysms.

Main Methods:

• Development of AView, an image-based computational tool for 3D morphological analysis of intracranial aneurysms.
• Validation of AView's accuracy using spheres of known dimensions and anatomical IA models.
• Comparison of manual 2D measurements with AView's 3D computer-assisted measurements on a retrospective clinical dataset.

Main Results:

• AView demonstrated high accuracy with an average error of 0.56% for size and 2.1% for volume measurements on model objects.
• Manual 2D measurements exhibited significant errors, averaging 22% for size and 30% for neck diameter in clinical cases.
• Computer-assisted analysis with AView improved measurement consistency by 62% for size and 82% for neck diameter among clinicians.

Conclusions:

• AView significantly enhances the accuracy and consistency of morphological analysis for intracranial aneurysms compared to traditional manual methods.
• The improved precision offered by AView is crucial for informed treatment decision-making, potentially reducing inappropriate strategies.
• A computer-assisted approach is essential for reliable morphological assessment of IAs, addressing the limitations of conventional 2D measurements.