Accuracy and interobserver reliability of three-dimensional rotational angiography versus mathematical models for volumetric measurement of intracranial aneurysms

Assessing accuracy and consistency in intracranial aneurysm sizing: human expertise vs. artificial intelligence

Intracranial aneurysms (IAs) are a common vascular pathology and are associated with a risk of rupture, which is often fatal. Aneurysm growth of more than 1 mm is considered a surrogate of rupture risk, therefore, this study presents a comprehensive analysis of intracranial aneurysm measurements utilizing a dataset comprising 358 IA from 248 computed tomography angiography (CTA) scans measured by four junior raters and one senior rater. The study explores the variability in sizing assessments by employing both human raters and an Artificial Intelligence (AI) system. Our findings reveal substantial inter- and intra-rater variability among junior raters, contrasting with the lower intra-rater variability observed in the senior rater. Standard deviations of all raters were above the threshold for IA growth (1 mm). Additionally, the study identifies a systemic bias, indicating a tendency for human experts to measure aneurysms smaller than the AI system. Our findings emphasize the challenges in human assessment while also showcasing the capacity of AI technology to improve the precision and reliability of intracranial aneurysm assessments, especially beneficial for junior raters. The potential of AI was particularly evident in the task of monitoring IA at various intervals, where the AI-based approach surpassed junior raters and achieved performance comparable to senior raters.

Interobserver variability in the characterization of giant intracranial aneurysms with special emphasis on aneurysm diameter and shape

BACKGROUND

The Giant Intracranial Aneurysm Registry is a multicenter observational trial exclusively focusing on giant intracranial aneurysms (GIA). As no data exist on the interobserver variability in the radiological description of GIA, there is some uncertainty concerning the reliability of the GIA characteristics included in the registry. We have therefore designed a study to test the interobserver variability in the description of the specific GIA characteristics that are examined in the GIA registry.

METHODS

Six different raters analyzed imaging of five GIA concerning GIA location, GIA size, GIA shape, GIA thrombosis, and the presence of perianeurysmal edema. Interobserver variability was examined using intraclass correlation and Fleiss' kappa analysis.

RESULTS

The intraclass correlation coefficient was 0.99 (95 % CI 0.97-1.0) for the largest GIA diameter and 0.98 (95 % CI 0.94-1.0) for the largest GIA diameter in an axial imaging slice. We found perfect interobserver agreement (Fleiss' kappa 1.00) in the characterization of GIA location and the presence of perianeurysmal edema and almost perfect interobserver agreement for GIA thrombosis (Fleiss' kappa 0.86, 95 % CI 0.63-1.00). Only moderate interobserver agreement was found in the description of GIA shape (Fleiss' kappa 0.50, 95 % CI 0.27-0.73).

CONCLUSIONS

While GIA size, location, thrombosis, and the presence of perianeurysmal edema showed excellent interobserver agreement, the description of GIA shape was achieved with only moderate agreement. Data on GIA shape in multicenter studies, like the GIA registry, should therefore be discussed with caution and potentially reassessed in a centralized fashion.

Influence of observers, threshold values, and measurement methods on volumetric analysis of cerebral aneurysms with three-dimensional rotational angiography

Introduction

Exact measurements of cerebral aneurysms are important in terms of their treatment. However, there is no definitive way to estimate the exact volume of an aneurysm. Our aim was to compare aneurysm volume measured under different conditions: threshold values, observers, and measurement methods.

Methods

40 aneurysms and 7 phantom models were included in the study. Three-dimensional rotational angiography was used, and volumes were compared based on the following factors: two methods (two-dimensional formula calculation and three-dimensional software measurement); three observers; and three threshold values for imaging. In addition, in the phantom models, measured volumes were compared with real volumes. The consistency of the volume measurements under different conditions was assessed using the intraclass correlation coefficient (ICC).

Results

The consistency of the measured volumes was excellent in both the patient aneurysms (ICC >0.98) and phantom models (ICC >0.95), irrespective of the influencing factors. Median volume differences were small for observers, threshold values, and methods. When the measured and real volumes of the phantom models were compared, the software measurement achieved the highest reproducibility for real volumes compared with the formula calculation (ICC=0.86–0.93 vs 0.71–0.80).

Conclusions

Measurement of aneurysm volume showed high agreement within each influencing factor, such as methods, observers, and threshold values.

In vivo embolization of lateral wall aneurysms in canines using the liquid-to-solid gelling PPODA-QT polymer system: 6-month pilot study

OBJECT

Over the past 20 years, endovascular embolization has become the preferred method of treating cerebral aneurysms. While there are many embolic devices on the market, none is ideal. In this study the authors investigated the use of a liquid-to-solid gelling polymer system-that is, poly(propylene glycol) diacrylate and pentaerythritol tetrakis (3-mercaptopropionate) (PPODA-QT)-to embolize in vivo aneurysms over a 6-month period.

METHODS

Experimental aneurysms were created in the carotid arteries of 9 canines. Aneurysms were embolized with the polymer only (PPODA-QT, 3 dogs), filled with PPODA-QT after placement of a "framing" platinum coil (coil+PPODA-QT, 3 dogs), or packed with platinum coils (coils only, 3 dogs). Aneurysm occlusion was angiographically monitored immediately and 6 months after embolization. After 6 months, the ostial regions of explanted aneurysms were assessed macroscopically and histologically.

RESULTS

All aneurysms showed 100% angiographic occlusion at 6 months, but turbulent blood flow was observed in 1 coils-only sample. Ostial regions of explanted coils-only aneurysms showed neointimal tissue surrounding individual coils but no continuous tissue layer over the aneurysm neck. All PPODA-QT aneurysms displayed smooth ostial surfaces, but 2 of 3 coil+PPODA-QT aneurysms showed polymer (unassociated with the coil) protruding into the vessel lumen, contributing to rough ostial surfaces. Neointimal tissue was present in PPODA-QT and coil+PPODA-QT aneurysms and covered smooth ostial surfaces more completely than in coils-only aneurysms.

CONCLUSIONS

This study compared neointimal tissue overgrowth in the ostium of experimental aneurysms embolized with PPODA-QT, PPODA-QT plus a framing coil, or coils alone. The coils-only and coil+PPODA-QT groups showed rough and discontinuous ostial surfaces, which hindered neointimal tissue coverage. The PPODA-QT aneurysms consistently produced smooth ostial surfaces that facilitated more complete neointimal tissue coverage over aneurysm necks.