Attention-Guided Autoencoder for Retrospective Physiological Anomaly Detection in ICU Vital Signs Using MIMIC-IV Demo Data
DOI:
https://doi.org/10.63158/journalisi.v8i5.1808Keywords:
Attention-Guided Autoencoder, ICU Vital Signs, MIMIC-IV Demo, Physiological Anomaly Screening, Reconstruction ErrorAbstract
This study evaluates threshold-defined physiological anomaly screening in retrospective vital-sign data from the MIMIC-IV Clinical Database Demo v2.2. Six vital signs were temporally aligned; Fahrenheit temperature values were converted to Celsius, physiologically implausible measurements were converted to missing values before within-stay forward/backward filling, and remaining incomplete rows were removed. After cleaning, the ICU stays were separated at stay level into 82 training, 28 validation, and 28 test stays. These partitions produced 9,210, 2,784, and 5,183 windows, respectively. Only normal windows from the training partition were used to fit the neural models, while labeled validation windows were used to choose the operating thresholds. Across five seeds, AGAE achieved accuracy 0.7838, precision 0.3417, recall 0.4954, specificity 0.8335, F1 0.4025, ROC-AUC 0.7709, and PR-AUC 0.3519. The threshold-independent results for AGAE and LSTM-AE were almost the same (ROC-AUC 0.7709 versus 0.7722; PR-AUC 0.3519 versus 0.3530). At the selected operating points, AGAE produced higher accuracy, precision, and specificity and fewer false positives; LSTM-AE produced higher recall and fewer false negatives. Taken together, these results do not show a consistent advantage for AGAE over LSTM-AE, and AGAE recall remained moderate. Prospective streaming and clinical validation are required before any real-time use.
Downloads
References
[1] F. Ali, S. El-Sappagh, S. M. R. Islam, A. Ali, M. Attique, M. Imran, and K.-S. Kwak, “An Intelligent Healthcare Monitoring Framework Using Wearable Sensors and Social Networking Data,” Future Gener. Comput. Syst., vol. 114, pp. 23–43, 2021, doi: 10.1016/j.future.2020.07.047.
[2] W. Nie, Y. Yu, C. Zhang, D. Song, L. Zhao, and Y. Bai, “Temporal-Spatial Correlation Attention Network for Clinical Data Analysis in Intensive Care Unit,” IEEE Trans. Biomed. Eng., vol. 71, no. 2, pp. 583–595, 2024, doi: 10.1109/TBME.2023.3309956.
[3] A. E. W. Johnson, L. Bulgarelli, L. Shen, et al., “MIMIC-IV, a Freely Accessible Electronic Health Record Dataset,” Sci. Data, vol. 10, Art. no. 1, 2023, doi: 10.1038/s41597-022-01899-x.
[4] K. A. Awan, S. Khan, K. Cengiz, H. Song, and I. Alrashdi, “MediNet: Self-Supervised Framework for Multimodal Analysis and Patient Care in IoMT,” IEEE Internet Things J., 2025, doi: 10.1109/JIOT.2025.3614207.
[5] K. Albanowski, K. J. Burdick, C. P. Bonafide, R. Kleinpell, and J. J. Schlesinger, “Ten Years Later, Alarm Fatigue Is Still a Safety Concern,” AACN Adv. Crit. Care, vol. 34, no. 3, pp. 189–197, 2023, doi: 10.4037/aacnacc2023662.
[6] Royal College of Physicians, National Early Warning Score (NEWS) 2: Standardising the Assessment of Acute-Illness Severity in the NHS, Updated Report of a Working Party. London, U.K.: Royal College of Physicians, 2017, ISBN: 978-1-86016-682-2.
[7] C. P. Subbe, M. Kruger, P. Rutherford, and L. Gemmel, “Validation of a Modified Early Warning Score in Medical Admissions,” QJM, vol. 94, no. 10, pp. 521–526, 2001, doi: 10.1093/qjmed/94.10.521.
[8] S. Choi, G. Kim, J. E. Kwon, and H. K. Lee, “Sequential Anomaly Detection for Continuous Prediction of Unexpected ICU Admission from Emergency Department,” IEEE Access, vol. 12, pp. 96304–96318, 2024, doi: 10.1109/ACCESS.2024.3426675.
[9] J. Fan, Z. Wang, H. Wu, D. Sun, J. Wu, and X. Lu, “An Adversarial Time–Frequency Reconstruction Network for Unsupervised Anomaly Detection,” Neural Netw., vol. 168, pp. 44–56, 2023, doi: 10.1016/j.neunet.2023.09.018.
[10] Z. Li and X. Zhang, “An Enhanced Autoencoder-Based Anomaly Detection Model for Time Series Data From Wearable Medical Devices,” IEEE J. Biomed. Health Inform., vol. 29, no. 10, pp. 7222–7233, 2025, doi: 10.1109/JBHI.2024.3434420.
[11] E. Karapalidou, N. Alexandris, E. Antoniou, S. Vologiannidis, J. Kalomiros, and D. Varsamis, “Implementation of a Sequence-to-Sequence Stacked Sparse Long Short-Term Memory Autoencoder for Anomaly Detection on Multivariate Timeseries Data of Industrial Blower Ball Bearing Units,” Sensors, vol. 23, no. 14, Art. no. 6502, 2023, doi: 10.3390/s23146502.
[12] S. Guan, Z. He, S. Ma, and M. Gao, “Conditional Normalizing Flow for Multivariate Time Series Anomaly Detection,” ISA Trans., vol. 143, pp. 231–243, 2023, doi: 10.1016/j.isatra.2023.09.002.
[13] W. Zhu, W. Li, E. R. Dorsey, and J. Luo, “Unsupervised Anomaly Detection by Densely Contrastive Learning for Time Series Data,” Neural Netw., vol. 168, pp. 450–458, 2023, doi: 10.1016/j.neunet.2023.09.038.
[14] J. Mi, Y. Zu, Z. Wang, and J. He, “ACDNet: Attention-Guided Collaborative Decision Network for Effective Medication Recommendation,” J. Biomed. Inform., vol. 149, Art. no. 104570, 2024, doi: 10.1016/j.jbi.2023.104570.
[15] R. Suzui, J. Natsume, M. Kawaguchi, Y. Ito, H. Yamamoto, T. Nakata, H. Kidokoro, Y. Takahashi, S. Tatsuki, and K. Fujiwara, “Prediction of Long-Term Prognosis in Infantile Epileptic Spasms Syndrome of Unknown Etiology Based on Hypsarrhythmia by Self-Attention Autoencoder,” IEEE J. Biomed. Health Inform., vol. 30, no. 7, pp. 5858–5868, 2026, doi: 10.1109/JBHI.2025.3647091.
[16] C. Wang, Z. Wang, H. Dong, S. Lauria, W. Liu, Y. Wang, F. Fadzil, and X. Liu, “Fusionformer: A Novel Adversarial Transformer Utilizing Fusion Attention for Multivariate Anomaly Detection,” IEEE Transactions on Neural Netw. and Learning Systems, vol. 36, no. 8, pp. 14479–14492, 2025, doi: 10.1109/TNNLS.2025.3542719.
[17] W. Hu, L. Cao, Q. Ruan, and Q. Wu, “Research on Anomaly Network Detection Based on Self-Attention Mechanism,” Sensors, vol. 23, no. 11, Art. no. 5059, 2023, doi: 10.3390/s23115059.
[18] Y. Chen, H. Zhang, Y. Wang, Y. Yang, X. Zhou, and Q. M. J. Wu, “MAMA Net: Multi-Scale Attention Memory Autoencoder Network for Anomaly Detection,” IEEE Trans. Med. Imaging, vol. 40, no. 3, pp. 1032–1041, 2021, doi: 10.1109/TMI.2020.3045295.
[19] A. Johnson, L. Bulgarelli, T. Pollard, S. Horng, L. A. Celi, and R. Mark, “MIMIC-IV Clinical Database Demo,” version 2.2, PhysioNet, 2023, doi: 10.13026/dp1f-ex47.
[20] S. Jain and B. C. Wallace, “Attention Is Not Explanation,” in Proc. 2019 Conf. North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), vol. 1, pp. 3543–3556, 2019, doi: 10.18653/v1/N19-1357.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Journal of Information Systems and Informatics
This work is licensed under a Creative Commons Attribution 4.0 International License.
Author's Declaration
- The Authors certify that they have read, understood, and agreed to the Journal of Information Systems and Informatics (JournalISI) submission guidelines, policies, and submission declaration. The submission has been prepared using the provided template.
- The Authors certify that all authors have approved the publication of this manuscript and that there is no conflict of interest.
- The Authors confirm that the manuscript is their original work, has not received prior publication, is not under consideration for publication elsewhere, and has not been previously published.
- The Authors confirm that all authors listed on the title page have contributed significantly to the work, have read the manuscript, attest to the validity and legitimacy of the data and its interpretation, and agree to its submission.
- The Authors confirm that the manuscript is not copied from or plagiarized from any other published work.
- The Authors declare that the manuscript will not be submitted for publication in any other journal or magazine until a decision is made by the journal editors.
- If the manuscript is finally accepted for publication, the Authors confirm that they will either proceed with publication immediately or withdraw the manuscript in accordance with the journal’s withdrawal policies.
- The Authors agree that, upon publication of the manuscript in this journal, they transfer copyright or assign exclusive rights to the publisher, including commercial rights
