AI Models for Cerebral Aneurysms Segmentation, Detection and Stability Prediction (AI-CARE)

Artificial Intelligence Applications for Cerebral Aneurysms Segmentation, Detection and Stability Prediction: a Stepwise, Multicenter Study

Aneurysmal subarachnoid hemorrhage (SAH) is one of the critical diseases that severely threaten human health, with a clinical mortality rate reaching as high as 30%. Early diagnosis and intervention before rupture are considered key to improving the prognosis of aneurysmal SAH. With the widespread clinical application of non-invasive cerebrovascular imaging techniques, such as CTA and MRA, the detection rate of unruptured intracranial aneurysms (UIAs) has significantly increased. However, addressing the growing demand for clinical cerebrovascular imaging diagnostics raises the challenge of improving diagnostic accuracy while alleviating the workload of diagnostic physicians. Furthermore, considering that not all detected UIAs will rupture, it is crucial to accurately identify high-risk aneurysms prone to rupture to avoid unnecessary overtreatment, which could lead to significant socioeconomic burdens and iatrogenic harm to patients.To meet this clinical need, researchers have developed an artificial intelligence (AI) algorithm to create software capable of automatically identifying intracranial aneurysms based on non-invasive vascular imaging data, enabling accurate diagnosis of aneurysms. To evaluate the clinical utility of this AI algorithm, a prospective, multicenter, registry study was proposed. Through long-term standardized and uniform non-invasive imaging follow-up, individualized imaging analysis profiles will be established. By correlating these profiles with aneurysm outcome events (growth or rupture), imaging features capable of accurately predicting aneurysm growth and rupture will be identified and analyzed. This approach is expected to enhance the accuracy of UIA diagnosis and enable risk stratification for unruptured intracranial aneurysms through the utilization of relevant data.

Study Overview

• Status: Recruiting
• Conditions: Subarachnoid Hemorrhage, Aneurysmal; Unruptured Cerebral Aneurysm; Magnetic Resonance Angiography; Artificial Intelligence (AI)

• Study Type: Observational
• Enrollment (Estimated): 10000

Contacts and Locations

• Study Contact: Name: Yueqi Zhu; Phone Number: +86-21-66301136; Email: zhuyueqi@hotmail.com

Study Locations

• China
  • Shanghai, China
    • Recruiting
    • Shanghai Sixth People's Hospital, Shanghai, 200023
    • Contact: Yueqi Zhu, MD; Phone Number: +86-21-66301136; Email: zhuyueqi@hotmail.com
    • Principal Investigator: Yueqi Zhu, MD

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

There is a preliminary diagnosis or symptoms indicating the presence or potential of a cerebral aneurysm.

Description

Inclusion Criteria:

1. Age ≥ 18 years;
2. Preliminary diagnosis or symptoms indicating the presence or potential presence of a cerebral aneurysm;
3. Undergoing a non-contrast head MRA or contrast-enhanced head/neck CTA;
4. The patient or their legal representative is able and willing to sign an informed consent form.

Exclusion Criteria:

1. Other intracranial vascular diseases: moyamoya disease, arteriovenous malformations, arteriovenous fistulas, arterial occlusions, and arterial dissections;
2. History of intracranial arterial interventions: stent placement, partial aneurysm coil treatment, etc.;
3. Severe allergy to contrast agents or absolute contraindications to iodine-based contrast agents;
4. Renal insufficiency with elevated serum creatinine (greater than twice the upper normal limit);
5. MRI contraindications: pacemakers, claustrophobia, etc.;
6. Diseases or conditions that affect the quality of CTA/MRA images;
7. Inability to complete the study due to psychiatric disorders, cognitive, or emotional disturbances.

Note: The CTA sub-study does not include exclusion criterion 5; the MRA sub-study does not include exclusion criteria 3 and 4.

Study Plan

How is the study designed?

Number of groups / cohorts

4

Cohorts and Interventions

Group / Cohort
Single MRA; The patient has undergone an MRA examination only once.
Multiple MRIs for one patient; The same patient underwent multiple MRAs.
MRA+DSA; The patient underwent both MRA and DSA within three months.
MRA+CTA; The patient underwent both MRA and CTA within three months.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Consistency between the artificial intelligence model and the manually annotated gold standard.	Consistency between the artificial intelligence model and the manually annotated gold standard.Intracranial aneurysms appear as saccular (most common), fusiform, or irregular localized vascular dilations on MRA 3D TOF（Time-of-Flight）.Aneurysms are classified into four main categories based on their diameter: small aneurysms, generally less than five millimeters; small aneurysms, with diameters ranging from 0.5 centimeters to 1.5 centimeters; large aneurysms, typically between 1.5 and 2.5 centimeters; and giant aneurysms, which exceed 2.5 centimeters.	1 month
Consistency between the artificial intelligence model and radiologists' image interpretations.	Consistency between the artificial intelligence model and radiologists' image interpretations.Intracranial aneurysms appear as saccular (most common), fusiform, or irregular localized vascular dilations on MRA 3D TOF（Time-of-Flight）.Aneurysms are classified into four main categories based on their diameter: small aneurysms, generally less than five millimeters; small aneurysms, with diameters ranging from 0.5 centimeters to 1.5 centimeters; large aneurysms, typically between 1.5 and 2.5 centimeters; and giant aneurysms, which exceed 2.5 centimeters.	1 month

Collaborators and Investigators

• Sponsor: Shanghai Jiao Tong University Affiliated Sixth People's Hospital

Study record dates

Study Major Dates

• Study Start (Actual): January 10, 2025
• Primary Completion (Estimated): June 30, 2026
• Study Completion (Estimated): June 30, 2027

Study Registration Dates

• First Submitted: December 25, 2024
• First Submitted That Met QC Criteria: January 4, 2025
• First Posted (Actual): January 9, 2025

Study Record Updates

• Last Update Posted (Actual): June 3, 2025
• Last Update Submitted That Met QC Criteria: May 31, 2025
• Last Verified: December 1, 2024

More Information

Terms related to this study

• Keywords: Artificial Intelligence; Subarachnoid Hemorrhage; Magnetic Resonance Angiography; Cerebral Aneurysms
• Additional Relevant MeSH Terms: Cerebrovascular Disorders; Brain Diseases; Central Nervous System Diseases; Nervous System Diseases; Vascular Diseases; Cardiovascular Diseases; Pathologic Processes; Intracranial Arterial Diseases; Intracranial Hemorrhages; Aneurysm; Hemorrhage; Subarachnoid Hemorrhage; Intracranial Aneurysm
• Other Study ID Numbers: AI-CARE

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No