Acute Stroke Treatment Eligibility with AI and Stroke Imaging


Determining if a patient can benefit from acute stroke treatments (endovascular thrombectomy and late window thrombolysis) relies on the use of advanced imaging modalities such as CT Perfusion (CTP) or MRI, which are not available at the majority of community hospitals and primary stroke centers.  Transferring all patients who could potentially need treatment to advanced-imaging capable centers is cost-prohibitive and impractical.


Investigators at UTHealth have addressed this problem by developing unique AI-machine learning methodologies to generate the types of information produced by CTP from a widely available imaging modality: CT angiogram.  This novel approach allows for screening of a vastly expanded population of patients for life-saving acute stroke treatments.


Key Benefits

-Ability to detect large vessel occlusion and predict ischemic core and penumbral volumes from CT angiogram with comparable accuracy to CTP studies.

-Potential to improve patient access to endovascular therapy

-Large dataset from Memorial Hermann (MH) Hospital System in Houston

-Live framework of immediate data transfer and analysis at multiple MH campuses

-Validated on multiple scanners, multiple hospitals, multiple cities

-Developed by interdisciplinary team combining stroke and machine learning expertise

-“One-stop-shop” with clinical and technical expertise for technology development and validation with multicenter data


Stage of Development

The tools have been validated in restrospective studies and is actively being tested prospectively at multiple Memorial Hermann Hospitals to optimize the clinical workflow.


Intellectual Property Status

Issued US Patent: US16/572,564

Available for licensing


Associated Publication

Sheth, S., et al. Machine Learning-Enabled Automated Determination of Acute Ischemic Core From Computed Tomography Angiography. Stroke. 2019 Nov; 50(11):3093-3100. doi: 10.1161/STROKEAHA.119.026189.

R. Abdelkhaleq et al., “Automated prediction of final infarct volume in patients with large-vessel occlusion acute ischemic stroke,” Neurosurgical Focus, vol. 51, no. 1, p. E13, Jul. 2021, doi: 10.3171/2021.4.FOCUS21134.

A. L. Czap et al., “Machine Learning Automated Detection of Large Vessel Occlusion From Mobile Stroke Unit Computed Tomography Angiography,” Stroke,2022; 53:1651–1656. doi: 10.1161/STROKEAHA.121.036091.

Giancardo, L., et al., “Segmentation of acute stroke infarct core using image-level labels on CT-angiography” Neuroimage Clin. 2023;37:103362. doi: 10.1016/j.nicl.2023.103362.


Lead Inventors

Dr. Sheth is a clinical faculty in the UTHealth McGovern Medical School Department of Neurology, who specialize in ischemic stroke.

Dr. Giancardo is an expert in machine learning, with a focus on applying machine learning to medical imaging, and is a faculty member in the UTHealth School of Biomedical Informatics.


UTHealth Ref. No.: 2018-0061



Patent Information:

The preceding is intended to be a non-confidential and limited description of a novel technology created at the University of Texas Health Science Center at Houston (UTHealth). This promotional material is not comprehensive in scope and should not replace company’s diligence in a thorough evaluation of the technology. Please contact the Office of Technology Management for more information regarding this technology.
Health IT
For Information, Contact:
Xiaoyan Wang
Technology Commercialization Analyst
University of Texas Health Science Center At Houston
Luca Giancardo
Sunil Sheth
© 2024. All Rights Reserved. Powered by Inteum