You are here

  • Home
  • Archive
  • Volume 15, Issue 6
  • Automated detection of arterial landmarks and vascular occlusions in patients with acute stroke receiving digital subtraction angiography using deep learning

Article Text

Article menu
Download PDFPDF
Ischemic stroke
Original research
Automated detection of arterial landmarks and vascular occlusions in patients with acute stroke receiving digital subtraction angiography using deep learning
  1. Jui Khankari1,
  2. Yannan Yu1,
  3. Jiahong Ouyang2,
  4. Ramy Hussein1,
  5. Huy M Do3,
  6. Jeremy J Heit4,
  7. Greg Zaharchuk1
  1. 1 Department of Radiology, Stanford University, Stanford, California, USA
  2. 2 Department of Electrical Engineering, Stanford University, Stanford, California, USA
  3. 3 Department of Radiology and Neurosurgery, Stanford University, Stanford, California, USA
  4. 4 Radiology, Neuroadiology and Neurointervention Division, Stanford University, Stanford, California, USA
  1. Correspondence to Dr Greg Zaharchuk, Department of Radiology, Stanford University, Stanford, CA 94305, USA; gregz{at}stanford.edu

Abstract

Background Digital subtraction angiography (DSA) is the gold-standard method of assessing arterial blood flow and blockages prior to endovascular thrombectomy.

Objective To detect anatomical features and arterial occlusions with DSA using artificial intelligence techniques.

Methods We included 82 patients with acute ischemic stroke who underwent DSA imaging and whose carotid terminus was visible in at least one run. Two neurointerventionalists labeled the carotid location (when visible) and vascular occlusions on 382 total individual DSA runs. For detecting the carotid terminus, positive and negative image patches (either containing or not containing the internal carotid artery terminus) were extracted in a 1:1 ratio. Two convolutional neural network architectures (ResNet-50 pretrained on ImageNet and ResNet-50 trained from scratch) were evaluated. Area under the curve (AUC) of the receiver operating characteristic and pixel distance from the ground truth were calculated. The same training and analysis methods were used for detecting arterial occlusions.

Results The ResNet-50 trained from scratch most accurately detected the carotid terminus (AUC 0.998 (95% CI 0.997 to 0.999), p<0.00001) and arterial occlusions (AUC 0.973 (95% CI 0.971 to 0.975), p<0.0001). Average pixel distances from ground truth for carotid terminus and occlusion localization were 63±45 and 98±84, corresponding to approximately 1.26±0.90 cm and 1.96±1.68 cm for a standard angiographic field-of-view.

Conclusion These results may serve as an unbiased standard for clinical stroke trials, as optimal standardization would be useful for core laboratories in endovascular thrombectomy studies, and also expedite decision-making during DSA-based procedures.

  • Stroke
  • Angiography

Data availability statement

Data are available upon reasonable request. Data are available upon request.

http://dx.doi.org/10.1136/neurintsurg-2021-018638

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Data availability statement

    Data are available upon reasonable request. Data are available upon request.

    View Full Text

    Footnotes

    • Twitter @JeremyHeitMDPHD

    • Contributors JK planned the machine learning experiments, conducted experiments and collected data, and reported the experiments in the paper. YY planned the scans, collected the data in the form of scans, and reported the data (in the form of scans) in the paper. JO planned the experiments, conducted the experiments, and reported the experiments in the paper. RH planned the experiments, conducted the experiments, and reported the experiments in the paper. HMD planned the scans, collected the data in the form of scans, and reported the data (in the form of scans) in the paper. JJH planned the scans, collected the data in the form of scans, and reported the data (in the form of scans) in the paper. GZ planned the scans and experiments, collected the data in the form of scans and experiments, reported the data (in the form of scans and machine learning experiments) in the paper, and is the guarantor.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests RH has received grant funding from NIH/NIA (P30 AG066515). JJH has received grant funding from NIH (PRECISE Study); minimal consulting fees from Medtronic, Inc and iSchemaView, and modest consulting fees from MicroVention, Inc; and participates on the Medical and Scientific Advisory Board of iSchemaView and on the clinical events committee of Balt and Vesalio. YY has received an AIMI Seed Grant. GZ has licenses for IP on deep learning, multiple patents on deep learning, and is the co-founder and has equity in Subtle Medical.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.