Comparative analysis and influencing factors of aneurysm wall characteristics based on artificial intelligence-assisted 3.0T MRI and 5.0T MRI high-resolution vascular wall imaging.

BACKGROUND

This study aimed to use artificial intelligence-assisted image processing methods to compare and analyze the effects of 3.0T and 5.0T high-resolution magnetic resonance vascular wall imaging (HR-VWI) in evaluating the wall characteristics of unruptured intracranial aneurysms, explore the influencing factors of aneurysm wall enhancement (AWE).

METHODS

Using 3.0T and 5.0T MRI scanners, cerebral vascular imaging was performed on 132 subjects and using deep convolutional neural networks to establish and validate an artificial intelligence model for automatic detection of aneurysms in MRI. According to the 5.0T HR-VWI results, patients were divided into an AWE group and a non-AWE group, and the differences in two-dimensional morphological parameters and clinical data between the two groups were further analyzed.

RESULTS

The detection rate of AWE in 5.0T HR-VWI was higher than that in 3.0T HR-VWI (P < 0.05). The incidence of AWE in patients with diabetes, smoking history, aneurysm diameter ≥ 7 mm and anterior communicating aneurysms was increased (P < 0.05). The incidence of wall enhancement in aneurysms with subcapsules and lobulated aneurysms was higher than that in aneurysms with regular morphology (P < 0.05). The aspect ratio and size ratio of aneurysms in the group with AWE were higher than those in the group without AWE (P < 0.05).

CONCLUSION

The incidence of AWE in diabetes, smoking history, anterior communicating artery aneurysms, aneurysms with diameter ≥ 7 mm, sacs, and lobulated aneurysms is increased. The AWE group showing increased aspect ratio and size ratio. 5.0T HR-VWI has a higher detection rate in identifying AWE. AI-assisted aneurysm diagnosis and measurement is expected to replace manual recognition and measurement to a certain extent.