Результаты исследований: Научные публикации в периодических изданиях › статья по материалам конференции › Рецензирование
Weakly supervised semantic segmentation of tomographic images in the diagnosis of stroke. / Dobshik, A. V.; Tulupov, A. A.; Berikov, V. B.
в: Journal of Physics: Conference Series, Том 2099, № 1, 012021, 13.12.2021.Результаты исследований: Научные публикации в периодических изданиях › статья по материалам конференции › Рецензирование
}
TY - JOUR
T1 - Weakly supervised semantic segmentation of tomographic images in the diagnosis of stroke
AU - Dobshik, A. V.
AU - Tulupov, A. A.
AU - Berikov, V. B.
N1 - Funding Information: The work was partly supported by RFBR grant 19-29-01175. Publisher Copyright: © 2021 Institute of Physics Publishing. All rights reserved.
PY - 2021/12/13
Y1 - 2021/12/13
N2 - This paper presents an automatic algorithm for the segmentation of areas affected by an acute stroke in the non-contrast computed tomography brain images. The proposed algorithm is designed for learning in a weakly supervised scenario when some images are labeled accurately, and some images are labeled inaccurately. Wrong labels appear as a result of inaccuracy made by a radiologist in the process of manual annotation of computed tomography images. We propose methods for solving the segmentation problem in the case of inaccurately labeled training data. We use the U-Net neural network architecture with several modifications. Experiments on real computed tomography scans show that the proposed methods increase the segmentation accuracy.
AB - This paper presents an automatic algorithm for the segmentation of areas affected by an acute stroke in the non-contrast computed tomography brain images. The proposed algorithm is designed for learning in a weakly supervised scenario when some images are labeled accurately, and some images are labeled inaccurately. Wrong labels appear as a result of inaccuracy made by a radiologist in the process of manual annotation of computed tomography images. We propose methods for solving the segmentation problem in the case of inaccurately labeled training data. We use the U-Net neural network architecture with several modifications. Experiments on real computed tomography scans show that the proposed methods increase the segmentation accuracy.
UR - http://www.scopus.com/inward/record.url?scp=85123727143&partnerID=8YFLogxK
U2 - 10.1088/1742-6596/2099/1/012021
DO - 10.1088/1742-6596/2099/1/012021
M3 - Conference article
AN - SCOPUS:85123727143
VL - 2099
JO - Journal of Physics: Conference Series
JF - Journal of Physics: Conference Series
SN - 1742-6588
IS - 1
M1 - 012021
T2 - International Conference on Marchuk Scientific Readings 2021, MSR 2021
Y2 - 4 October 2021 through 8 October 2021
ER -
ID: 35377098