Feasibility of a Deep Learning-based Algorithm for Non-invasive Assessment of Vulnerable Coronary Plaque

Not yet recruiting

• Conditions: Coronary Artery Disease; Chronic Coronary Syndrome; Acute Coronary Syndrome; Stable Angina
• Interventions: Other: Deep Learning-based Vulnerable Plaque Detection and Assessment Tool

• Registration Number: NCT06186336
• Lead Sponsor: GE Healthcare

Brief Summary

The primary objective of this study is to assess the accuracy in terms of sensitivity, specificity, negative and positive predicted values of the DL-based algorithm with respect to correct identification of the plaque and associated vulnerability grade.

Detailed Description

Data collected in this study will be used for technology development, scientific evaluation, education, and regulatory submissions for future products. This is a pre-market, open label, prospective, non-randomized clinical research study conducted at one site in Italy. The product being researched is the Deep Learning-based (DL) algorithm for non-invasive detection of vulnerable coronary plaque.

Study Timeline

• Study First Submitted: 2023-12-15
• Study First Submitted QC: 2023-12-15
• Study First Posted: 2024-01-02
• Status Verified: 2024-11
• Study Start: 2025-01
• Last Update Submitted: 2025-01-06
• Last Update Posted: 2025-01-07
• Primary Completion: 2025-06
• Study Completion: 2025-06

Recruitment & Eligibility

• Status: NOT_YET_RECRUITING
• Sex: All
• Target Recruitment: 200

Inclusion Criteria

1. Patients referred for a clinically indicated CCTA and ICA with OCT imaging examinations;
2. Diagnosis of chronic coronary syndrome, known CAD, or stable angina; AND,
3. Patients with ACS that may undergo a CCTA and not refer directly to the Cath lab for revascularization procedures.

Exclusion Criteria

1. Contradictions to contrast;
2. Contraindications for beta blocker;
3. BMI >30;
4. High heart rate ≥75 BPM;
5. Atrial Fibrillation;
6. Arrythmia or irregular heartbeats;
7. Any prior coronary revascularization;
8. Presence of pacemaker or implantable cardioverter defibrillator; OR,
9. Patients with TAVI/TAVR.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Arm && Interventions

Group	Intervention	Description
DL-Based Vulnerable Plaque Detection and Assessment Tool	Deep Learning-based Vulnerable Plaque Detection and Assessment Tool	Enrolled subjects will receive a clinically indicated CCTA and ICA with OCT within 10 days. At least two qualified CCTA radiologists will independently review and annotate the coronary plaques in the CCTA using their local post-processing tools. At least two trained OCT readers will review and annotate the coronary plaques in the ICA/OCT using their local post-processing tools. The original de-identified CCTA data will be inputted into the Vulnerable Plaque Detection and Assessment Tool. The tool will perform the automatic identification of the plaque location and characteristics. These results will be compared to assess the algorithm's performance.

Primary Outcome Measures

Name	Time	Method
Data Collection	Through study completion, an average of 1 year	Number of subjects with raw CCTA and ICA with OCT data

Secondary Outcome Measures

Name	Time	Method
Accuracy of Tool	Through study completion, an average of 1 year	Sensitivity, specificity, negative, and positive predicted values of the DL-based algorithm with respect to correct identification of the plaque and associated vulnerability grade in comparison to qualified readers
Safety Events	Through study completion, an average of 1 year	Number of AEs, SAEs, and product issues

Trial Locations

Locations (1)

Ospedale Di Voghera, Azienda Socio-Sanitaria Territoriale di Pavia; 🇮🇹 Pavia, Lombardy, Italy