Stratification of carotid atheromatous plaque using interpretable deep learning methods on B-mode ultrasound images

被引:4
作者
Ganitidis, Theofanis [1 ]
Athanasiou, Maria [1 ]
Dalakleidi, Kalliopi [1 ]
Melanitis, Nikos [1 ]
Golemati, Spyretta [1 ,2 ]
Nikita, Konstantina S. [1 ]
机构
[1] Natl Tech Univ Athens NTUA, Biomed Simulat & Imaging BIOSIM Lab, 9 Iroon Polytech Str, Zografos 15780, Greece
[2] Natl & Kapodistrian Univ Athens, Med Sch, Athens, Greece
来源
2021 43RD ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE & BIOLOGY SOCIETY (EMBC) | 2021年
关键词
Carotid; image analysis; ultrasound; deep learning; medical imaging; interpretability; explainable AI; ATHEROSCLEROSIS; CLASSIFICATION;
D O I
10.1109/EMBC46164.2021.9630402
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
Carotid atherosclerosis is the major cause of ischemic stroke resulting in significant rates of mortality and disability annually. Early diagnosis of such cases is of great importance, since it enables clinicians to apply a more effective treatment strategy. This paper introduces an interpretable classification approach of carotid ultrasound images for the risk assessment and stratification of patients with carotid atheromatous plaque. To address the highly imbalanced distribution of patients between the symptomatic and asymptomatic classes (16 vs 58, respectively), an ensemble learning scheme based on a sub-sampling approach was applied along with a two-phase, cost-sensitive strategy of learning, that uses the original and a resampled data set. Convolutional Neural Networks (CNNs) were utilized for building the primary models of the ensemble. A six-layer deep CNN was used to automatically extract features from the images, followed by a classification stage of two fully connected layers. The obtained results (Area Under the ROC Curve (AUC): 73% sensitivity: 75% specificity: 70%) indicate that the proposed approach achieved acceptable discrimination performance. Finally, interpretability methods were applied on the model's predictions in order to reveal insights on the model's decision process as well as to enable the identification of novel image biomarkers for the stratification of patients with carotid atheromatous plaque.
引用
收藏
页码:3902 / 3905
页数:4
相关论文
共 50 条
  • [31] A deep learning framework to assess the feasibility of localizing prostate cancer on b-mode transrectal ultrasound images
    Jahanandish, Hassan
    Vesal, Sulaiman
    Bhattacharya, Indrani
    Li, Cynthia Xinran
    Fan, Richard E.
    Sonn, Geoffrey A.
    Rusu, Mirabela
    MEDICAL IMAGING 2024: ULTRASONIC IMAGING AND TOMOGRAPHY, 2024, 12932
  • [32] Lumen Segmentation and Motion Estimation in B-Mode and Contrast-Enhanced Ultrasound Images of the Carotid Artery in Patients With Atherosclerotic Plaque
    Carvalho, Diego D. B.
    Akkus, Zeynettin
    van den Oord, Stijn C. H.
    Schinkel, Arend F. L.
    van der Steen, Antonius F. W.
    Niessen, Wiro J.
    Bosch, Johan G.
    Klein, Stefan
    IEEE TRANSACTIONS ON MEDICAL IMAGING, 2015, 34 (04) : 983 - 993
  • [33] Automatic segmentation of carotid B-mode images using fuzzy classification
    Rocha, Rui
    Silva, Jorge
    Campilho, Aurelio
    MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING, 2012, 50 (05) : 533 - 545
  • [34] Automatic segmentation of carotid B-mode images using fuzzy classification
    Rui Rocha
    Jorge Silva
    Aurélio Campilho
    Medical & Biological Engineering & Computing, 2012, 50 : 533 - 545
  • [35] Synthetic Elastography from B-Mode ultrasound through Deep Learning
    Wildeboer, R. R.
    van Sloun, R. J. G.
    Mannaerts, Christophe K.
    Salomon, G.
    Wijkstra, H.
    Mischi, M.
    2019 IEEE INTERNATIONAL ULTRASONICS SYMPOSIUM (IUS), 2019, : 108 - 110
  • [36] A novel automatic algorithm for the segmentation of the lumen of the carotid artery in ultrasound B-mode images
    Santos, Andre Miguel F.
    dos Santos, Rosa Maria
    Castro, Pedro Miguel A. C.
    Azevedo, Elsa
    Sousa, Luisa
    Tavares, Joao Manuel R. S.
    EXPERT SYSTEMS WITH APPLICATIONS, 2013, 40 (16) : 6570 - 6579
  • [37] Carotid plaque score and ischemic stroke risk stratification through a combination of B-mode and contrast-enhanced ultrasound in patients with low and intermediate carotid stenosis
    Li, Fang
    Gu, Shi-Yao
    Zhang, Lu-Ni
    Chen, Jing
    Yao, Ming-Hua
    Wu, Ting-Ting
    Ma, Ji
    Jia, Cai-Xia
    Wu, Rong
    FRONTIERS IN CARDIOVASCULAR MEDICINE, 2023, 10
  • [38] B-Flow and B-Mode Ultrasound Imaging in Carotid Fibromuscular Dysplasia
    Cutler, Joshua J.
    Campo, Nelly
    Koch, Sebastian
    JOURNAL OF NEUROIMAGING, 2018, 28 (03) : 269 - 272
  • [39] B-mode ultrasound to elastography synthesis using multiscale learning
    Dai, Fei
    Li, Yifang
    Zhu, Yunkai
    Li, Boyi
    Shi, Qinzhen
    Chen, Yaqing
    Ta, Dean
    ULTRASONICS, 2024, 138
  • [40] Combined B-mode and Nakagami Images for Improved Discrimination of Breast Masses using Deep Learning
    Muhtadi, Sabiq
    Haque, Syed Tousiful
    Gallippi, Caterina M.
    2022 IEEE INTERNATIONAL ULTRASONICS SYMPOSIUM (IEEE IUS), 2022,