Go Green! Reusing Brain Monitoring Data Containing Missing Values: A Feasibility Study with Traumatic Brain Injury Patie
Background: Despite the wealth of information carried, periodic brain monitoring data are often incomplete with a significant amount of missing values. Incomplete monitoring data are usually discarded to ensure purity of data. However, this approach leads
- PDF / 453,450 Bytes
- 9 Pages / 595.28 x 790.87 pts Page_size
- 119 Downloads / 189 Views
Abstract Background: Despite the wealth of information carried, periodic brain monitoring data are often incomplete with a significant amount of missing values. Incomplete monitoring data are usually discarded to ensure purity of data. However, this approach leads to the loss of statistical power, potentially biased study and a great waste of resources. Thus, we propose to reuse incomplete brain monitoring data by imputing the missing values – a green solution! To support our proposal, we have conducted a feasibility study to investigate the reusability of incomplete brain monitoring data based on the estimated imputation error. Materials and Methods: Seventy-seven patients, who underwent invasive monitoring of ICP, MAP, PbtO2 and brain temperature (BTemp) for more than 24 consecutive hours and were connected to a bedside computerized system, were selected for the study. In the feasibility study, the imputation error is experimentally assessed with simulated missing values and 17 state-of-the-art predictive methods. A framework is developed for neuroclinicians and neurosurgeons to determine the best re-usage strategy and predictive methods based on our feasibility study.
M. Feng (), L.Y. Loy, F. Zhang, Z. Zhang, K. Vellaisamy, P.L. Chin, and C. Guan Institute for Infocomm Research, A*STAR, 1 Fusionopolis Way, #21-01, Connexis (South), Singapore, Singapore e-mail: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected] L. Shen Unit of Biostatistics, Yong Loo Lin School of Medicine, National University of Singapore, Block 1E Kent Ridge Road, Singapore, Singapore e-mail: [email protected] N.K.K. King, K.K. Lee, and B.T. Ang Department of Neurosurgery, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore, Singapore e-mail: [email protected]; [email protected]; [email protected]
Results/Conclusion: The monitoring data of MAP and BTemp are more reliable for reuse than ICP and PbtO2; and, for ICP and PbtO2 data, a more cautious re-usage strategy should be employed. We also observe that, for the scenarios tested, the lazy learning method, K-STAR, and the tree-based method, M5P, are consistently 2 of the best among the 17 predictive methods investigated in this study. Keywords Traumatic brain injury • Brain monitoring • Intracranial pressure • Missing values/data • Data re-usage • Feasibility study
Introduction Background Traumatic brain injury (TBI) [5] constitutes a large portion of the workload of neuro-critical care, and it is a major cause of death and disabilities worldwide [20]. The major challenge in TBI patient treatment is that: primary traumatic brain damage is often compounded by secondary pathophysiological insults that occur after trauma, during the patient’s stay in the intensive care unit [10]. Many studies [13, 22] have demonstrated that secondary insults have significant effects on the mortality and recovery rates of TBI patients. Secondary insults can
Data Loading...
