Detecting the left atrial appendage in CT localizers using deep learning

Abstract Patients with cardioembolic stroke often undergo CT of the left atrial appendage (LAA), for example, to determine whether thrombi are present in the LAA. To guide the imaging process, technologists first perform a localizer scan, which is a preliminary image used to identify the region of i...

Full description

Saved in:
Bibliographic Details
Main Authors: Aydin Demircioğlu, Denise Bos, Anton S. Quinsten, Lale Umutlu, Oliver Bruder, Michael Forsting, Kai Nassenstein
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-99701-6
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849314723960455168
author Aydin Demircioğlu
Denise Bos
Anton S. Quinsten
Lale Umutlu
Oliver Bruder
Michael Forsting
Kai Nassenstein
author_facet Aydin Demircioğlu
Denise Bos
Anton S. Quinsten
Lale Umutlu
Oliver Bruder
Michael Forsting
Kai Nassenstein
author_sort Aydin Demircioğlu
collection DOAJ
description Abstract Patients with cardioembolic stroke often undergo CT of the left atrial appendage (LAA), for example, to determine whether thrombi are present in the LAA. To guide the imaging process, technologists first perform a localizer scan, which is a preliminary image used to identify the region of interest. However, the lack of well-defined landmarks makes accurate delimitation of the LAA in localizers difficult and often requires whole-heart scans, increasing radiation exposure and cancer risk. This study aims to automate LAA delimitation in CT localizers using deep learning. Four commonly used deep networks (VariFocalNet, Cascade-R-CNN, Task-aligned One-stage Object Detection Network, YOLO v11) were trained to predict the LAA boundaries on a cohort of 1253 localizers, collected retrospectively from a single center. The best-performing network in terms of delimitation accuracy was then evaluated on an internal test cohort of 368 patients, and on an external test cohort of 309 patients. The VariFocalNet performed best, achieving LAA delimitations with high accuracy (97.8% and 96.8%; Dice coefficients: 90.4% and 90.0%) and near-perfect clinical utility (99.8% and 99.3%). Compared to whole-heart scanning, the network-based delimitation reduced the radiation exposure by more than 50% (5.33 ± 6.42 mSv vs. 11.35 ± 8.17 mSv in the internal cohort, 4.39 ± 4.23 mSv vs. 10.09 ± 8.0 mSv in the external cohort). This study demonstrates that a deep learning network can accurately delimit the LAA in the localizer, leading to more accurate CT scans of the LAA, thereby significantly reducing radiation exposure to the patient compared to whole-heart scanning.
format Article
id doaj-art-b902bd279ada44af87fa46ff5e83f7a4
institution Kabale University
issn 2045-2322
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-b902bd279ada44af87fa46ff5e83f7a42025-08-20T03:52:23ZengNature PortfolioScientific Reports2045-23222025-05-0115111110.1038/s41598-025-99701-6Detecting the left atrial appendage in CT localizers using deep learningAydin Demircioğlu0Denise Bos1Anton S. Quinsten2Lale Umutlu3Oliver Bruder4Michael Forsting5Kai Nassenstein6Institute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital EssenInstitute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital EssenInstitute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital EssenInstitute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital EssenDepartment of Cardiology and Angiology, Contilia Heart and Vascular Center, Elisabeth-Krankenhaus EssenInstitute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital EssenInstitute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital EssenAbstract Patients with cardioembolic stroke often undergo CT of the left atrial appendage (LAA), for example, to determine whether thrombi are present in the LAA. To guide the imaging process, technologists first perform a localizer scan, which is a preliminary image used to identify the region of interest. However, the lack of well-defined landmarks makes accurate delimitation of the LAA in localizers difficult and often requires whole-heart scans, increasing radiation exposure and cancer risk. This study aims to automate LAA delimitation in CT localizers using deep learning. Four commonly used deep networks (VariFocalNet, Cascade-R-CNN, Task-aligned One-stage Object Detection Network, YOLO v11) were trained to predict the LAA boundaries on a cohort of 1253 localizers, collected retrospectively from a single center. The best-performing network in terms of delimitation accuracy was then evaluated on an internal test cohort of 368 patients, and on an external test cohort of 309 patients. The VariFocalNet performed best, achieving LAA delimitations with high accuracy (97.8% and 96.8%; Dice coefficients: 90.4% and 90.0%) and near-perfect clinical utility (99.8% and 99.3%). Compared to whole-heart scanning, the network-based delimitation reduced the radiation exposure by more than 50% (5.33 ± 6.42 mSv vs. 11.35 ± 8.17 mSv in the internal cohort, 4.39 ± 4.23 mSv vs. 10.09 ± 8.0 mSv in the external cohort). This study demonstrates that a deep learning network can accurately delimit the LAA in the localizer, leading to more accurate CT scans of the LAA, thereby significantly reducing radiation exposure to the patient compared to whole-heart scanning.https://doi.org/10.1038/s41598-025-99701-6Radiation safetyCT localizerCoronary CT angiographyHeartDeep learning
spellingShingle Aydin Demircioğlu
Denise Bos
Anton S. Quinsten
Lale Umutlu
Oliver Bruder
Michael Forsting
Kai Nassenstein
Detecting the left atrial appendage in CT localizers using deep learning
Scientific Reports
Radiation safety
CT localizer
Coronary CT angiography
Heart
Deep learning
title Detecting the left atrial appendage in CT localizers using deep learning
title_full Detecting the left atrial appendage in CT localizers using deep learning
title_fullStr Detecting the left atrial appendage in CT localizers using deep learning
title_full_unstemmed Detecting the left atrial appendage in CT localizers using deep learning
title_short Detecting the left atrial appendage in CT localizers using deep learning
title_sort detecting the left atrial appendage in ct localizers using deep learning
topic Radiation safety
CT localizer
Coronary CT angiography
Heart
Deep learning
url https://doi.org/10.1038/s41598-025-99701-6
work_keys_str_mv AT aydindemircioglu detectingtheleftatrialappendageinctlocalizersusingdeeplearning
AT denisebos detectingtheleftatrialappendageinctlocalizersusingdeeplearning
AT antonsquinsten detectingtheleftatrialappendageinctlocalizersusingdeeplearning
AT laleumutlu detectingtheleftatrialappendageinctlocalizersusingdeeplearning
AT oliverbruder detectingtheleftatrialappendageinctlocalizersusingdeeplearning
AT michaelforsting detectingtheleftatrialappendageinctlocalizersusingdeeplearning
AT kainassenstein detectingtheleftatrialappendageinctlocalizersusingdeeplearning