Recent advances in computer and sensor technology have enabled considerable progress in augmenting and improving the skills of medical practitioners. This study proposes a new data-driven approach to support training of endotracheal intubation (ETI), a procedure performed to prevent suffocation of an unconscious person, by analyzing hand motion and heterogeneous sensor features. ETI is one of the most important requisite skills in emergency or intensive care that must be trained and practiced regularly to maintain proficiency. Currently, however, ETI training is assessed by human supervisors who may make inconsistent assessments. This study aims at developing an effective and efficient automated assessment system that can analyze ETI skills and classify the trainee into an experienced or a novice provider immediately after training. To make the system more available and affordable, first we investigate the feasibility of utilizing only hand motion data as a determining factor of ETI proficiency. In order to further improve the classification performance, we find the best feature combination from heterogeneous sensors in addition to hand motion features. We extract 18 features from hand motion in time and frequency domain, and also 12, 4 and 8 features for tongue force, incisor force and head angle, respectively. Subsequently, feature selection algorithms, including correlation-based feature selection, Relief, and mRMR, are applied to the features to identify an ideal feature set for classification. An artificial neural network (ANN) classifier is trained by using each selected feature set respectively. Experimental results show that the ANN classifier with five hand motion features selected by the correlation-based algorithm achieves the highest accuracy of 91.17% among the results obtained by individual sensors. Also, an ANN with a heterogeneous feature set, which consists of six features extracted from hand motion, incisor force and head angle, achieves the highest accuracy of 96.40% for ETI assessment. This study corroborates that a simple assessment system based on a small number of features obtained from hand motion and multi-sensor fusion can be effective in assisting ETI training.
|Advisor:||Ko, Hoo Sang, Cho, Sohyung|
|School:||Southern Illinois University at Edwardsville|
|School Location:||United States -- Illinois|
|Source:||MAI 81/3(E), Masters Abstracts International|
|Keywords:||Assessment, Classification, Data mining, Endotracheal intubation|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be