TY - GEN

T1 - Segmentation of 3D Non-Contrast CT Brain Images Using Transformer Neural Networks

AU - Kirillov, Kirill

AU - Mikhailapov, Denis

AU - Tulupov, Andrey

AU - Berikov, Vladimir

N1 - The work was partly supported by the State Contract of the Sobolev Institute of Mathematics, Project No. FWNF-2022-0015.

PY - 2024/11/26

Y1 - 2024/11/26

N2 - We propose a method for semantic segmentation of 3D non-contrast computed tomography brain images of acute ischemic stroke using transformer neural networks. To improve the segmentation quality of lesion areas, the pre-processing methods were implemented. The 3D Swin UNETR model is employed for segmentation, which is based on the attention mechanism. The sum of DICE loss and Focal loss are used to train the model, and DICE score as well as sensitivity and precision is utilized to evaluate the quality of model's predictions. The model was trained and tested using cross-validation on real images of patients at the International Tomography Center SB RAS. Research and comparison of the performance of the model and its analogues was carried out. The proposed algorithm demonstrates 30% greater DICE metric in comparison with the analogous 3D U-Net model. The main feature of the 3D Swin UNETR model is the increase in false positives and the decrease in false negatives compared to 3D U-Net.

AB - We propose a method for semantic segmentation of 3D non-contrast computed tomography brain images of acute ischemic stroke using transformer neural networks. To improve the segmentation quality of lesion areas, the pre-processing methods were implemented. The 3D Swin UNETR model is employed for segmentation, which is based on the attention mechanism. The sum of DICE loss and Focal loss are used to train the model, and DICE score as well as sensitivity and precision is utilized to evaluate the quality of model's predictions. The model was trained and tested using cross-validation on real images of patients at the International Tomography Center SB RAS. Research and comparison of the performance of the model and its analogues was carried out. The proposed algorithm demonstrates 30% greater DICE metric in comparison with the analogous 3D U-Net model. The main feature of the 3D Swin UNETR model is the increase in false positives and the decrease in false negatives compared to 3D U-Net.

KW - ischemic stroke

KW - neural networks

KW - non-contrast CT

KW - semantic segmentation

KW - swin-transformer

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85212142837&origin=inward&txGid=d1ca057b68ae387e31b9e901039e847e

UR - https://www.mendeley.com/catalogue/f81dbba8-b606-3870-bb54-8b290b99c752/

U2 - 10.1109/SIBIRCON63777.2024.10758478

DO - 10.1109/SIBIRCON63777.2024.10758478

M3 - Conference contribution

SN - 9798331532024

T3 - 2024 IEEE International Multi-Conference on Engineering, Computer and Information Sciences, SIBIRCON 2024

SP - 173

EP - 177

BT - 2024 IEEE International Multi-Conference on Engineering, Computer and Information Sciences, SIBIRCON 2024

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2024 IEEE International Multi-Conference on Engineering, Computer and Information Sciences

Y2 - 30 September 2024 through 2 November 2024

ER -