Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets
Anomaly detection (AD) is a challenging problem in computer vision. Particularly in the field of medical imaging, AD poses even more challenges due to a number of reasons, including insufficient availability of ground truth (annotated) data. In recent years, AD models based on generative adversarial...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10043696/ |
| _version_ | 1797894937375670272 |
|---|---|
| author | Marzieh Esmaeili Amirhosein Toosi Arash Roshanpoor Vahid Changizi Marjan Ghazisaeedi Arman Rahmim Mohammad Sabokrou |
| author_facet | Marzieh Esmaeili Amirhosein Toosi Arash Roshanpoor Vahid Changizi Marjan Ghazisaeedi Arman Rahmim Mohammad Sabokrou |
| author_sort | Marzieh Esmaeili |
| collection | DOAJ |
| description | Anomaly detection (AD) is a challenging problem in computer vision. Particularly in the field of medical imaging, AD poses even more challenges due to a number of reasons, including insufficient availability of ground truth (annotated) data. In recent years, AD models based on generative adversarial networks (GANs) have made significant progress. However, their effectiveness in biomedical imaging remains underexplored. In this paper, we present an overview of using GANs for AD, as well as an investigation of state-of-the-art GAN-based AD methods for biomedical imaging and the challenges encountered in detail. We have also specifically investigated the advantages and limitations of AD methods on medical image datasets, conducting experiments using 3 AD methods on 7 medical imaging datasets from different modalities and organs/tissues. Given the highly different findings achieved across these experiments, we further analyzed the results from both data-centric and model-centric points of view. The results showed that none of the methods had a reliable performance for detecting abnormalities in medical images. Factors such as the number of training samples, the subtlety of the anomaly, and the dispersion of the anomaly in the images are among the phenomena that highly impact the performance of the AD models. The obtained results were highly variable (AUC: 0.475-0.991; Sensitivity: 0.17-0.98; Specificity: 0.14-0.97). In addition, we provide recommendations for the deployment of AD models in medical imaging and foresee important research directions. |
| first_indexed | 2024-04-10T07:18:29Z |
| format | Article |
| id | doaj.art-5d1205963ab149fe8fea25a72efe6ad8 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-10T07:18:29Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-5d1205963ab149fe8fea25a72efe6ad82023-02-25T00:01:47ZengIEEEIEEE Access2169-35362023-01-0111179061792110.1109/ACCESS.2023.324474110043696Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image DatasetsMarzieh Esmaeili0https://orcid.org/0000-0002-3559-4852Amirhosein Toosi1Arash Roshanpoor2Vahid Changizi3Marjan Ghazisaeedi4https://orcid.org/0000-0002-2400-209XArman Rahmim5https://orcid.org/0000-0002-9980-2403Mohammad Sabokrou6https://orcid.org/0000-0002-9409-2799Department of Health Information Management, School of Allied Medical Sciences, Tehran University of Medical Sciences, Tehran, IranDepartment of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, CanadaDepartment of Computer, Yadegar-e-Imam Khomeini, Janat-Abad Branch, Islamic Azad University, Tehran, IranDepartment of Radiology and Radiotherapy Technology, School of Allied Health Sciences, Tehran University of Medical Sciences, Tehran, IranDepartment of Health Information Management, School of Allied Medical Sciences, Tehran University of Medical Sciences, Tehran, IranDepartment of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, CanadaSchool of Computer Science, Institute for Research in Fundamental Sciences (IPM), Tehran, IranAnomaly detection (AD) is a challenging problem in computer vision. Particularly in the field of medical imaging, AD poses even more challenges due to a number of reasons, including insufficient availability of ground truth (annotated) data. In recent years, AD models based on generative adversarial networks (GANs) have made significant progress. However, their effectiveness in biomedical imaging remains underexplored. In this paper, we present an overview of using GANs for AD, as well as an investigation of state-of-the-art GAN-based AD methods for biomedical imaging and the challenges encountered in detail. We have also specifically investigated the advantages and limitations of AD methods on medical image datasets, conducting experiments using 3 AD methods on 7 medical imaging datasets from different modalities and organs/tissues. Given the highly different findings achieved across these experiments, we further analyzed the results from both data-centric and model-centric points of view. The results showed that none of the methods had a reliable performance for detecting abnormalities in medical images. Factors such as the number of training samples, the subtlety of the anomaly, and the dispersion of the anomaly in the images are among the phenomena that highly impact the performance of the AD models. The obtained results were highly variable (AUC: 0.475-0.991; Sensitivity: 0.17-0.98; Specificity: 0.14-0.97). In addition, we provide recommendations for the deployment of AD models in medical imaging and foresee important research directions.https://ieeexplore.ieee.org/document/10043696/Anomaly detectionartificial intelligencemachine learningdeep learningunsupervised anomaly detectiongenerative adversarial networks |
| spellingShingle | Marzieh Esmaeili Amirhosein Toosi Arash Roshanpoor Vahid Changizi Marjan Ghazisaeedi Arman Rahmim Mohammad Sabokrou Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets IEEE Access Anomaly detection artificial intelligence machine learning deep learning unsupervised anomaly detection generative adversarial networks |
| title | Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets |
| title_full | Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets |
| title_fullStr | Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets |
| title_full_unstemmed | Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets |
| title_short | Generative Adversarial Networks for Anomaly Detection in Biomedical Imaging: A Study on Seven Medical Image Datasets |
| title_sort | generative adversarial networks for anomaly detection in biomedical imaging a study on seven medical image datasets |
| topic | Anomaly detection artificial intelligence machine learning deep learning unsupervised anomaly detection generative adversarial networks |
| url | https://ieeexplore.ieee.org/document/10043696/ |
| work_keys_str_mv | AT marziehesmaeili generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets AT amirhoseintoosi generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets AT arashroshanpoor generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets AT vahidchangizi generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets AT marjanghazisaeedi generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets AT armanrahmim generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets AT mohammadsabokrou generativeadversarialnetworksforanomalydetectioninbiomedicalimagingastudyonsevenmedicalimagedatasets |