Analysis and Application of Regression Models to ICU Patient Monitoring

Abstract

When a patient is admitted to the hospital, they undergo monitoring and a variety of data is collected. The continuous stream of clinical and analytical data provides a wealth of information. However, circumstances may arise where certain medical test values are not acquired. Therefore, this chapter aims to examine the medical test values obtained from ICU patients with COVID-19 and utilize regression models to predict any missing values. By applying these regression models, we can obtain predictions of the medical test values, enabling us to retrieve missing data and provide medical staff with an additional tool and more information to combat COVID-19. This study introduces a novel approach to tackle the issue of missing medical data in ICU patients. Unlike previous studies that have focused on retrieving medical values from a single test, our model enables the retrieval of values from multiple medical tests simultaneously. The study evaluated the performance of the proposed model using several metrics, including RMSE, MAPE, and MAE. Applying the proposed model yielded a prediction of retrieved medical evidence values with an RMSE of 4.81, MAPE of 0.071, and MAE of 3.26. Additionally, the article presented different scenarios in which the model demonstrated higher accuracy. The model developed in this study has the potential to be adapted and used for the diagnosis of future infectious diseases with similar characteristics to COVID-19. Providing accurate and reliable predictions of medical test values can be a valuable tool for clinicians and medical staff to make timely and informed decisions regarding patient care. Ultimately, this can lead to better management and treatment of infectious diseases such as COVID-19, which can improve patient outcomes.