Study on the Performance of a Machine Learning Algorithm Recognizing and Triaging Large Vessel Occlusions Using Non-contrast CT Scans (SMART-LVO)

Methinks Software

Status: Not yet enrolling

Conditions: Brain Ischemia; Stroke Hemorrhagic; Stroke, Acute; Stroke, Ischemic

Treatments: Device: AI software for Stroke

Study type: Observational

Funder types: Industry; Other

Identifiers

NCT06216457

SMART-LVO Registry

Details and patient eligibility

About

The goal of this prospective observational study is to assess the effectiveness and performance of Methinks AI stroke imaging software platform in acute Code Stroke patients, and as a comparator to study sites utilizing existing AI imaging stroke platforms.

The main question[s] it aims to answer is:

• Performance of and outcomes associated with the use of the Methinks AI stroke imaging medical device in real-world clinical practice.

Full description

This is a prospective observational medical device clinical trial that aims to assess the diagnostic performance of Methinks Investigational Device on detecting LVO and ICH on plain NCCT and CTA, in consecutive acute Code Stroke activations at the emergency department.

Demographic, Clinical and Imaging variables will be collected from stroke onset through Day 7 or Discharge, whichever is earlier. Methinks results will be collected exclusively for research purposes and no clinical decisions will be made based on the output. Diagnostic performance of NCCT-LVO Methinks output will be assessed using DSA and/or CTA as ground truth. Presence of ICH will be assessed using an independent neuroradiologist final reading. The Clinical impact on workflow will be retrospectively estimated based on the reduction of time to decision in both transfer patients and the ones directly arriving to the emergency department (ED)

Informed consent forms (ICF) will be waived, as approved by the IRB.

Enrollment: 5,000 estimated patients

Sex: All

Ages: 22+ years old

Volunteers: No Healthy Volunteers

Inclusion criteria

• Subjects with acute neurological symptoms suggestive of Stroke, ischemic or hemorrhagic ("Code Stroke"), regardless of presence or absence of final Stroke diagnosis.
• Subjects aged 22 years or older.
• NCCT performed for Code Stroke screening.
• If the patient does not have a hemorrhage, a CTA (head and neck, performed for Code Stroke screening) should also be performed.
• NCCT and CTA have been processed by Methinks AI software.

Exclusion criteria

● Subject imaging does not meet Image Acquisition and DICOM Tag Requirements

Trial design

5,000 participants in 3 patient groups

Group 1: LVO

Description:

Subjects with Methinks NCCT algorithm detecting ICA, M1, and M2 occlusions.

Treatment:

Device: AI software for Stroke

Group 2: ICH

Description:

Subjects with Methinks NCCT algorithm detecting ICHs.

Treatment:

Device: AI software for Stroke

Group 3: Controls

Description:

Subjects with final diagnosis confirmed not to be Stroke (i.e. stroke mimic) or any type of stroke not determined to be: LVO (ICA, M1 or M2) or ICH.

Treatment:

Device: AI software for Stroke