Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 May 20;13(1):36.
doi: 10.1007/s13755-025-00352-8. eCollection 2025 Dec.

Cerebral ischemia detection using deep learning techniques

Rafael Pastor-Vargas #  1 , Cristina Antón-Munárriz  2 , Juan M Haut #  3 , Antonio Robles-Gómez #  1 , Mercedes E Paoletti #  3 , José Alberto Benítez-Andrades #  4
Affiliations

Cerebral ischemia detection using deep learning techniques

Rafael Pastor-Vargas et al. Health Inf Sci Syst. .

Abstract

Cerebrovascular accident (CVA), commonly known as stroke, stands as a significant contributor to contemporary mortality and morbidity rates, often leading to lasting disabilities. Early identification is crucial in mitigating its impact and reducing mortality. Non-contrast computed tomography (NCCT) remains the primary diagnostic tool in stroke emergencies due to its speed, accessibility, and cost-effectiveness. NCCT enables the exclusion of hemorrhage and directs attention to ischemic causes resulting from arterial flow obstruction. Quantification of NCCT findings employs the Alberta Stroke Program Early Computed Tomography Score (ASPECTS), which evaluates affected brain structures. This study seeks to identify early alterations in NCCT density in patients with stroke symptoms using a binary classifier distinguishing NCCT scans with and without stroke. To achieve this, various well-known deep learning architectures, namely VGG3D, ResNet3D, and DenseNet3D, validated in the ImageNet challenges, are implemented with 3D images covering the entire brain volume. The training results of these networks are presented, wherein diverse parameters are examined for optimal performance. The DenseNet3D network emerges as the most effective model, attaining a training set accuracy of 98% and a test set accuracy of 95%. The aim is to alert medical professionals to potential stroke cases in their early stages based on NCCT findings displaying altered density patterns.

Keywords: Cerebral ischemia; Computed tomography; Deep learning; Ictus dataset; Transfer learning.

PubMed Disclaimer

Conflict of interest statement

Conflict of interestAll authors declare that there is no conflict of interest.

Figures

Fig. 1
Fig. 1
ASPECTS areas: Insula (I), Lenticular (L), Caudate (C), Capsule Internal (CI) and cortical areas: M1, M2, M3, M4, M5 and M6
Fig. 2
Fig. 2
CT-image characteristics
Fig. 3
Fig. 3
View of ASPECTS punctuation’s areas, detailing an ischemia in M2 and L areas
Fig. 4
Fig. 4
Comparison between original and skull deleted CTs
Fig. 5
Fig. 5
DenseNet results, version 1.1.1
Fig. 6
Fig. 6
DenseNet results, version 1.2.1
Fig. 7
Fig. 7
DenseNet results, version 1.3
Fig. 8
Fig. 8
DenseNet 3D results, version 2.2
Fig. 9
Fig. 9
ResNet 3D results, version 1.0
Fig. 10
Fig. 10
ResNet 3D results, version 1.1
Fig. 11
Fig. 11
ResNet 3D results, version 2.0
Fig. 12
Fig. 12
ResNet 3D results, version 2.1
Fig. 13
Fig. 13
VGG 3D results: version 3.0
Fig. 14
Fig. 14
VGG 3D results, version 3.1
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Antón-Munárriz C, Pastor-Vargas R, Haut JM, et al (2023) Detection of cerebral ischaemia using transfer learning techniques. In: 2023 IEEE 36th international symposium on computer-based medical systems (CBMS), pp 589–59. 10.1109/CBMS58004.2023.00284
    1. Oman O, Makela T, Salli E, et al. 3d convolutional neural networks applied to ct angiography in the detection of acute ischemic stroke. Eur Radiol Exp. 2019;3(1):8. - PMC - PubMed
    1. Ramos MM, Giadas TC. Evaluación vascular en el código ictus: papel de la angio-tomografía computarizada. Radiologia. 2015;57(2):156–66. - PubMed
    1. Barber PA, Demchuk AM, Zhang J, et al. Validity and reliability of a quantitative computed tomography score in predicting outcome of hyperacute stroke before thrombolytic therapy. Lancet. 2000;355(9216):1670–4. - PubMed
    1. Lisowska A, O’Neil A, Dilys V, et al (2017) Context-aware convolutional neural networks for stroke sign detection in non-contrast ct scans. In: Annual conference on medical image understanding and analysis, Springer, pp 494–505

LinkOut - more resources