Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

AI for Radiology: A Primer Part II. Interacting with AI Results.

Radiology·2026
Same author

Domain Generalization Mitigates Scanner-Induced Domain Shift in Medical Imaging.

Journal of imaging informatics in medicine·2026
Same author

Detection of hypodense hepatic and renal lesions on abdominal CT reconstructed with deep learning image reconstruction technique in patients with large body habitus: A multi-reader study.

Abdominal radiology (New York)·2026
Same author

Radiology Reimagined: Interoperability and Lessons Learned from the Imaging AI in Practice Demonstration.

Radiology·2026
Same author

Glymphatic dysfunction is unlikely to cause idiopathic intracranial hypertension.

Headache·2026
Same author

Weak supervision of H&E slides reveals systems-level biology and functional states that govern therapeutic resistance.

bioRxiv : the preprint server for biology·2026

Related Experiment Video

Updated: Aug 15, 2025

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.7K

Head CT deep learning model is highly accurate for early infarct estimation.

Romane Gauriau1, Bernardo C Bizzo2,3,4, Donnella S Comeau1

  • 1Data Science Office, Mass General Brigham, 100 Cambridge St, Suite 1303, Boston, MA, 02114, USA.

Scientific Reports
|January 5, 2023
PubMed
Summary

A new deep learning model significantly improves early acute infarct detection on non-contrast head CT scans. This AI tool enhances stroke diagnosis accuracy, outperforming expert neuroradiologists in identifying suspected early ischemic strokes.

More Related Videos

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

48.2K
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.9K

Related Experiment Videos

Last Updated: Aug 15, 2025

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.7K
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

48.2K
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.9K

Area of Science:

  • Radiology
  • Artificial Intelligence
  • Neurology

Background:

  • Non-contrast head CT (NCCT) has limited sensitivity for detecting early acute ischemic strokes within the first 3-6 hours.
  • Accurate and timely identification of acute infarcts is crucial for effective stroke treatment, particularly thrombectomy.

Purpose of the Study:

  • To develop and validate a deep learning (DL) model for detecting and delineating early acute infarcts on NCCT.
  • To compare the DL model's performance against expert neuroradiologists in identifying infarcts.

Main Methods:

  • A DL model was trained on 3566 NCCT/MRI patient pairs, using diffusion MRI as the ground truth.
  • The model was evaluated on a test set of 150 NCCT scans from patients eligible for thrombectomy.
  • Performance was further assessed on an expanded dataset of 364 NCCT scans with diverse infarct locations and volumes.

Main Results:

  • The DL model demonstrated superior sensitivity (96%) compared to expert neuroradiologists (61-66%) in identifying infarcts on the initial test set.
  • Infarct volume estimates from the DL model showed strong correlation with diffusion MRI (r² > 0.98).
  • On the larger, heterogeneous dataset, the model achieved 97% sensitivity and 99% specificity for detecting infarcts exceeding 70 mL.

Conclusions:

  • Deep learning models can significantly enhance the detection of early acute ischemic infarcts on NCCT.
  • The developed DL model shows high accuracy and outperforms human experts, offering a promising tool for stroke diagnosis and patient selection for thrombectomy.