Learning disentangled representations for explainable chest X-ray classification using Dirichlet VAEs

被引:0
|
作者
Harkness, Rachael [1 ,2 ]
Frangi, Alejandro F. [1 ,2 ]
Zucker, Kieran [2 ,3 ]
Ravikumar, Nishant [1 ,2 ]
机构
[1] Univ Leeds, Sch Comp, CISTIB Ctr Computat Imaging & Simulat Technol Bio, Leeds, W Yorkshire, England
[2] Univ Leeds, Leeds, W Yorkshire, England
[3] LIMR Leeds Inst Med Res, Leeds, W Yorkshire, England
来源
MEDICAL IMAGING 2023 | 2023年 / 12464卷
关键词
Multi-label classification; chest X-ray images; variational autoencoders; Dirichlet distribution; disentanglement; explainability;
D O I
10.1117/12.2654345
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This study explores the use of the Dirichlet Variational Autoencoder (DirVAE) for learning disentangled latent representations of chest X-ray (CXR) images. Our working hypothesis is that distributional sparsity, as facilitated by the Dirichlet prior, will encourage disentangled feature learning for the complex task of multi-label classification of CXR images. The DirVAE is trained using CXR images from the CheXpert database, and the predictive capacity of multi-modal latent representations learned by DirVAE models is investigated through implementation of an auxiliary multi-label classification task, with a view to enforce separation of latent factors according to class-specific features. The predictive performance and explainability of the latent space learned using the DirVAE were quantitatively and qualitatively assessed, respectively, and compared with a standard Gaussian prior-VAE (GVAE). We introduce a new approach for explainable multi-label classification in which we conduct gradient-guided latent traversals for each class of interest. Study findings indicate that the DirVAE is able to disentangle latent factors into class-specific visual features, a property not afforded by the GVAE, and achieve a marginal increase in predictive performance relative to GVAE. We generate visual examples to show that our explainability method, when applied to the trained DirVAE, is able to highlight regions in CXR images that are clinically relevant to the class(es) of interest and additionally, can identify cases where classification relies on spurious feature correlations.
引用
收藏
页数:13
相关论文
共 50 条
  • [1] Learning Interpretable Disentangled Representations Using Adversarial VAEs
    Sarhan, Mhd Hasan
    Eslami, Abouzar
    Navab, Nassir
    Albarqouni, Shadi
    DOMAIN ADAPTATION AND REPRESENTATION TRANSFER AND MEDICAL IMAGE LEARNING WITH LESS LABELS AND IMPERFECT DATA, DART 2019, MIL3ID 2019, 2019, 11795 : 37 - 44
  • [2] Explainable Knowledge Distillation for On-Device Chest X-Ray Classification
    Termritthikun, Chakkrit
    Umer, Ayaz
    Suwanwimolkul, Suwichaya
    Xia, Feng
    Lee, Ivan
    IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS, 2024, 21 (04) : 846 - 856
  • [3] Chest X-Ray classification using transfer learning on multi GPU
    Ponomaryov, Volodymyr, I
    Almaraz-Damian, Jose A.
    Reyes-Reyes, Rogelio
    Cruz-Ramos, Clara
    REAL-TIME IMAGE PROCESSING AND DEEP LEARNING 2021, 2021, 11736
  • [4] Explainable Lung Disease Classification from Chest X-Ray Images Utilizing Deep Learning and XAI
    Ifty, Tanzina Taher
    Shafin, Saleh Ahmed
    Shahriar, Shoeb Mohammad
    Towhid, Tashfia
    2024 IEEE 3RD INTERNATIONAL CONFERENCE ON COMPUTING AND MACHINE INTELLIGENCE, ICMI 2024, 2024,
  • [5] DEEP LEARNING CLASSIFICATION OF CHEST X-RAY IMAGES
    Majdi, Mohammad S.
    Salman, Khalil N.
    Morris, Michael F.
    Merchant, Nirav C.
    Rodriguez, Jeffrey J.
    2020 IEEE SOUTHWEST SYMPOSIUM ON IMAGE ANALYSIS AND INTERPRETATION (SSIAI 2020), 2020, : 116 - 119
  • [6] Evaluating Local Explainable AI Techniques for the Classification of Chest X-Ray Images
    Sciacca, Enrico
    Estatico, Claudio
    Verda, Damiano
    Ferrari, Enrico
    EXPLAINABLE ARTIFICIAL INTELLIGENCE, XAI 2024, PT IV, 2024, 2156 : 66 - 83
  • [7] Explainable COVID-19 Three Classes Severity Classification Using Chest X-Ray Images
    Thon, P. L.
    Than, J. C. M.
    Kassim, R. M.
    Yunus, A.
    Noor, N. M.
    Then, P.
    2022 IEEE-EMBS CONFERENCE ON BIOMEDICAL ENGINEERING AND SCIENCES, IECBES, 2022, : 312 - 317
  • [8] Continual Learning for Domain Adaptation in Chest X-ray Classification
    Lenga, Matthias
    Schulz, Heinrich
    Saalbach, Axel
    MEDICAL IMAGING WITH DEEP LEARNING, VOL 121, 2020, 121 : 413 - 423
  • [9] Chest X-ray and CT Scan Classification using Ensemble Learning through Transfer Learning
    Siddiqui, Salman Ahmad
    Fatima, Neda
    Ahmad, Anwar
    EAI ENDORSED TRANSACTIONS ON SCALABLE INFORMATION SYSTEMS, 2022, 9 (06)
  • [10] Disentangled Contrastive Learning From Synthetic Matching Pairs for Targeted Chest X-Ray Generation
    Kim, Euyoung
    Lee, Soochahn
    Mu Lee, Kyoung
    IEEE ACCESS, 2025, 13 : 15453 - 15468