TY - JOUR

T1 - Adjusting U-Net for the aortic abdominal aneurysm CT segmentation case

AU - Epifanov, Rostislav Uirevich

AU - Nikitin, Nikita Aleksandrovich

AU - Rabtsun, Artem Aleksandrovich

AU - Kurdyukov, Leonid Nicolaevich

AU - Karpenko, Andrey Anatolievich

AU - Mullyadzhanov, Rustam Ilhamovich

N1 - The work is supported by the Russian Science Foundation grant No. 21-15-00091.

PY - 2024/5

Y1 - 2024/5

N2 - In this paper, we address the issue of developing of a convolutional neural network for the problem of aneurysm segmentation into three classes and of exploring ways for improving the quality of final segmentation masks. As a result of our study, macro dice score for classes of interest reaches 83.12 % ± 4.27 %. We explored different augmentation styles and showed the importance of applying intensity augmentation style to improve segmentation algorithm robustness in conditions of clinical data diversity. Augmentation with spatial and insensitive styles increase macro dice score up to 3 %. The comparison of various inference mode indicate that combination of overlapping inference and segmentation window enlargement ameliorate macro dice up to 1.4 %. Overall improvement of the quality of segmentation masks by macro dice score amounted up to 6 % using combination of data-based augmentation style and advanced inference technique.

AB - In this paper, we address the issue of developing of a convolutional neural network for the problem of aneurysm segmentation into three classes and of exploring ways for improving the quality of final segmentation masks. As a result of our study, macro dice score for classes of interest reaches 83.12 % ± 4.27 %. We explored different augmentation styles and showed the importance of applying intensity augmentation style to improve segmentation algorithm robustness in conditions of clinical data diversity. Augmentation with spatial and insensitive styles increase macro dice score up to 3 %. The comparison of various inference mode indicate that combination of overlapping inference and segmentation window enlargement ameliorate macro dice up to 1.4 %. Overall improvement of the quality of segmentation masks by macro dice score amounted up to 6 % using combination of data-based augmentation style and advanced inference technique.

KW - abdominal aortic aneurysm

KW - calcifications

KW - neural network

KW - semantic segmentation

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85208415775&origin=inward&txGid=afe2390a9ac6c0ca96613a26218ea913

UR - https://www.mendeley.com/catalogue/4a33435e-ca58-3c1f-b75f-1f2a49504e8d/

U2 - 10.18287/2412-6179-CO-1338

DO - 10.18287/2412-6179-CO-1338

M3 - Article

VL - 48

SP - 418

EP - 424

JO - Computer Optics

JF - Computer Optics

SN - 0134-2452

IS - 3

ER -