Evaluation of predictive models of aneurysm focal growth and bleb development using machine learning techniques

Radiomic profiling of high-risk aneurysms with blebs: an exploratory study

BACKGROUND

Blebs significantly increase rupture risk of intracranial aneurysms. Radiomic analysis offers a robust characterization of the aneurysm wall. However, the unique radiomic profile of various compartments, including blebs, remains unexplored. Likewise, the correlation between these imaging markers and fluid/mechanical metrics is yet to be investigated. To address this, we analyzed the radiomic features (RFs) of bleb-containing aneurysms and their relationship with wall tension and shear stress metrics, aiming to enhance risk assessment.

METHODS

Aneurysms were imaged using high-resolution magnetic resonance imaging (MRI). A T1 and a T1 after contrast (T1+Gd) sequences were acquired. 3D models of aneurysm bodies and blebs were generated, and RFs were extracted. Aneurysms with and without blebs were matched based on location and size for analysis. Univariate regression models and Spearman's correlations were used to establish associations between bleb-dependent RFs and mechanical/fluid dynamics metrics.

RESULTS

Eighteen aneurysms with blebs were identified. Fifty-five RFs were significantly different between blebs and body within the same aneurysms. Of these RFs, 9% (5/55) were first-order, and 91% (50/55) were second-order features. After aneurysms with and without blebs were matched for location and size, five RFs 5% (5/93) were significantly different. Forty-one out of the 55 RFs different between bleb and body sac of the primary aneurysm were moderately and strongly correlated with mechanical and fluid dynamics metrics.

CONCLUSION

Aneurysm blebs exhibit distinct radiomic profiles compared with the main body of the aneurysm sac. The variability in bleb wall characteristics may arise from differing mechanical stresses and localized hemodynamics. Leveraging radiomic profiling could help identify regions with a heightened risk of rupture.

Global trend in research of intracranial aneurysm management with artificial intelligence technology: a bibliometric analysis

Background

The use of artificial intelligence (AI) technology has been growing in the management of intracranial aneurysms (IAs). This study aims to conduct a bibliometric analysis of researches on intracranial aneurysm management with artificial intelligence technology (IAMWAIT) to gain insights into global research trends and potential future directions.

Methods

A comprehensive search of articles and reviews related to IAMWAIT, published from January 1, 1900 to July 20, 2023, was conducted using the Web of Science Core Collection (WoWCC).Visualizations of the bibliometric analysis were generated utilizing WPS Office, Scimago Graphica, VOSviewer, CiteSpace, and R.

Results

A total of 277 papers were included in the study. China emerged as the most prolific country in terms of publications, institutions, cooperating countries, and prolific authors. The United States garnered the highest number of total citations, institutions with the highest citations/H index, cooperating countries (n=9), and 3 of the top 10 cited papers. Both the total number of papers and the citation count exhibited a positive and significant correlation with the gross domestic product (GDP) of countries. The journal with the highest publication frequency was Frontiers in Neurology, while Stroke recorded the highest number of citations, H-index, and impact factor (IF). Areas of primary interest in IAMWAIT, leveraging AI technology, included rupture risk assessment/prediction, computer-assisted diagnosis, outcome prediction, hemodynamics, and laboratory research of IAs.

Conclusions

IAMWAIT is an active area of research that has undergone rapid development in recent years. Future endeavors should focus on broader application of AI algorithms in various sub-fields of IAMWAIT to better suit the real world.