DenseNet121 sebagai Feature Extractor pada Denoising Autoencoder untuk Deteksi Anomali Unsupervised Citra X-Ray Dada
- Raihan Muhammar Zikra Universitas Islam Negeri Sultan Syarif Kasim, Riau, Indonesia
- Febi Yanto Universitas Islam Negeri Sultan Syarif Kasim, Riau, Indonesia
-
Benny Sukma Negara
*
Universitas Islam Negeri Sultan Syarif Kasim, Riau,
Indonesia
- Surya Agustian Universitas Islam Negeri Sultan Syarif Kasim, Riau, Indonesia
- Reski Mai Candra Universitas Islam Negeri Sultan Syarif Kasim, Riau, Indonesia (*) Corresponding Author
Abstract
Anomaly detection in chest X-ray images remains a challenge in medical imaging, as supervised learning approaches require large amounts of labeled data that are difficult and costly to annotate. This study proposes an unsupervised learning anomaly detection system that integrates a pretrained DenseNet121 as a feature extractor with a Denoising Autoencoder (DAE), so that training requires only normal images without anomaly annotations. The model was trained on normal images and tested on COVID-19 and pneumonia images to evaluate its anomaly detection capability based on reconstruction error relative to an optimal threshold. Evaluation was conducted on the Covid19-Pneumonia-Normal Chest X-Ray Images dataset comprising 5,228 images, comparing the performance of DenseNet121 and ResNet50 as feature extractors across three latent dimension configurations. The DenseNet121 configuration with a latent dimension of 128 achieved the highest overall performance on most metrics, namely 91% accuracy, 90.75% sensitivity, 72% Macro F1-Score, and a validation loss (MSE) of 0.0952 on 3,608 test images, although its AUC (0.9526) and specificity (87.36%) were not consistently the highest among all tested configurations. These results demonstrate that using DenseNet121 as a feature extractor improves the DAE's ability to distinguish normal from anomalous lung images, suggesting its potential as an efficient preliminary screening approach under conditions of limited labeled data.
Downloads
References
Alrayes, F. S., Zakariah, M., Amin, S. U., Iqbal Khan, Z., & Helal, M. (2024). Intrusion Detection in IoT Systems Using Denoising Autoencoder. IEEE Access, 12, 122401–122425. https://doi.org/10.1109/ACCESS.2024.3451726
Alvarado, W., Agrawal, V., Li, W. S., Dravid, V. P., Backman, V., de Pablo, J. J., & Ferguson, A. L. (2023). Denoising Autoencoder Trained on Simulation-Derived Structures for Noise Reduction in Chromatin Scanning Transmission Electron Microscopy. ACS Central Science, 9(6), 1200–1212. https://doi.org/10.1021/acscentsci.3c00178
Chicco, D., & Jurman, G. (2023). The Matthews correlation coefficient (MCC) should replace the ROC AUC as the standard metric for assessing binary classification. BioData Mining, 16(1), 4. https://doi.org/10.1186/s13040-023-00322-4
FAROOQ, Y., & SAVAŞ, S. (2024). Noise Removal from the Image Using Convolutional Neural Networks-Based Denoising Auto Encoder. Journal of Emerging Computer Technologies, 3(1), 21–28. https://doi.org/10.57020/ject.1390428
He, J., Xu, Q., Jiang, Y., Wang, Z., & Huang, Q. (2024). ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection. Proceedings of the AAAI Conference on Artificial Intelligence, 38(8), 8481–8489. https://doi.org/10.1609/aaai.v38i8.28691
Kascenas, A., Pugeault, N., & O’Neil, A. Q. (2022). Denoising Autoencoders for Unsupervised Anomaly Detection in Brain MRI. Proceedings of Machine Learning Research, 172, 653–664.
Kim, S.-M., & Soo Kim, Y. (2024). Enhancing Sound-Based Anomaly Detection Using Deep Denoising Autoencoder. IEEE Access, 12, 84323–84332. https://doi.org/10.1109/ACCESS.2024.3414435
Liu, T., Li, B., Du, X., Jiang, B., Geng, L., Wang, F., & Zhao, Z. (2025). Simple and effective Frequency-aware Image Restoration for industrial visual anomaly detection. Advanced Engineering Informatics, 64, 103064. https://doi.org/10.1016/j.aei.2024.103064
Mujkic, E., Philipsen, M. P., Moeslund, T. B., Christiansen, M. P., & Ravn, O. (2022). Anomaly Detection for Agricultural Vehicles Using Autoencoders. Sensors, 22(10), 3608. https://doi.org/10.3390/s22103608
Nawaz, A., Khan, S. S., & Ahmad, A. (2024). Ensemble of Autoencoders for Anomaly Detection in Biomedical Data: A Narrative Review. IEEE Access, 12, 17273–17289. https://doi.org/10.1109/ACCESS.2024.3360691
Park, H., Jeong, J., Oh, K.-W., & Kim, H. (2023). Autoencoder-Based Recommender System Exploiting Natural Noise Removal. IEEE Access, 11, 30609–30618. https://doi.org/10.1109/ACCESS.2023.3262026
Patil, R., & Bhosale, S. (2022). Medical Image Denoising Techniques: A Review. International Journal on Engineering, Science and Technology, 4(1), 21–33. https://doi.org/10.46328/ijonest.76
Raghuvamsi, Y., Teeparthi, K., & Kosana, V. (2023). Denoising autoencoder based topology identification in distribution systems with missing measurements. International Journal of Electrical Power & Energy Systems, 154, 109464. https://doi.org/10.1016/j.ijepes.2023.109464
Rajput, G., Agrawal, S., Biyani, K., & Vishvakarma, S. K. (2022). Early breast cancer diagnosis using cogent activation function‐based deep learning implementation on screened mammograms. International Journal of Imaging Systems and Technology, 32(4), 1101–1118. https://doi.org/10.1002/ima.22701
Sachin Kumar. (2022). Covid19-Pneumonia-Normal Chest X-Ray Images. Mendeley Data. https://doi.org/10.17632/dvntn9yhd2.1
Scarpiniti, M., Sarv Ahrabi, S., Baccarelli, E., Piazzo, L., & Momenzadeh, A. (2022). A novel unsupervised approach based on the hidden features of Deep Denoising Autoencoders for COVID-19 disease detection. Expert Systems with Applications, 192, 116366. https://doi.org/10.1016/j.eswa.2021.116366
Siddiqi, R., & Javaid, S. (2024). Deep Learning for Pneumonia Detection in Chest X-ray Images: A Comprehensive Survey. Journal of Imaging, 10(8), 176. https://doi.org/10.3390/jimaging10080176
Thiele, C., & Hirschfeld, G. (2021). cutpointr : Improved Estimation and Validation of Optimal Cutpoints in R. Journal of Statistical Software, 98(11). https://doi.org/10.18637/jss.v098.i11
Torabi, H., Mirtaheri, S. L., & Greco, S. (2023). Practical autoencoder based anomaly detection by using vector reconstruction error. Cybersecurity, 6(1), 1. https://doi.org/10.1186/s42400-022-00134-9
Yang, H., Aydi, W., Innab, N., Ghoneim, M. E., & Ferrara, M. (2024). Classification of cervical cancer using Dense CapsNet with Seg-UNet and denoising autoencoders. Scientific Reports, 14(1), 31764. https://doi.org/10.1038/s41598-024-82489-2
Zhou, S., He, Z., Chen, X., & Chang, W. (2024). An Anomaly Detection Method for UAV Based on Wavelet Decomposition and Stacked Denoising Autoencoder. Aerospace, 11(5), 393. https://doi.org/10.3390/aerospace11050393
Bila bermanfaat silahkan share artikel ini
Berikan Komentar Anda terhadap artikel DenseNet121 sebagai Feature Extractor pada Denoising Autoencoder untuk Deteksi Anomali Unsupervised Citra X-Ray Dada
Copyright (c) 2026 Raihan Muhammar Zikra, Febi Yanto, Benny Sukma Negara, Surya Agustian, Reski Mai Candra
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under Creative Commons Attribution 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (Refer to The Effect of Open Access).
