Investigating the Use of LSTM and Time-Series Analysis in Medical Equipment Failure Prediction

• Author(s): Precious Ejiba ; Philip Nwaga ; Ugochi Awuoki ; Odera Ohazurike
• Paper ID: 1706462
• Page: 1-9
• Published Date: 30-10-2024
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 8 Issue 5 November-2024

Abstract

Ensuring the reliability of medical equipment is crucial for uninterrupted healthcare delivery, as unexpected failures can lead to significant operational disruptions and affect patient safety. This study explores the potential of Long Short-Term Memory (LSTM) neural networks for predicting equipment failures using historical time-series data. The methodology involved pre-processing failure data, implementing a sliding window approach for feature engineering, and training an LSTM model to capture patterns indicative of potential equipment failures. The results showed that the LSTM model could recognize trends in historical data; however, the model’s loss curve exhibited variability, suggesting limitations in achieving consistent accuracy. This fluctuation points to areas where the model’s robustness could be improved, particularly by enhancing data quality and exploring additional predictive features. While the findings support the application of LSTM networks in predictive maintenance, further research is recommended to validate the model in real-world healthcare environments and optimize it for practical implementation. This study contributes to the growing body of research on machine learning in predictive maintenance, underscoring both the promise and challenges of applying advanced neural networks to healthcare equipment management.

Keywords

Long short-term memory (LSTM), Predictive maintenance (PdM), Medical equipment, Machine learning, Time-series analysis.

Citations

IRE Journals:
Precious Ejiba , Philip Nwaga , Ugochi Awuoki , Odera Ohazurike "Investigating the Use of LSTM and Time-Series Analysis in Medical Equipment Failure Prediction" Iconic Research And Engineering Journals Volume 8 Issue 5 2024 Page 1-9

IEEE:
Precious Ejiba , Philip Nwaga , Ugochi Awuoki , Odera Ohazurike "Investigating the Use of LSTM and Time-Series Analysis in Medical Equipment Failure Prediction" Iconic Research And Engineering Journals, 8(5)