Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2014
A portable multi-channel wireless NIRS device for muscle activity real-time monitoring.
Near-infrared spectroscopy (NIRS) is a relative new technology in monitoring muscle oxygenation and hemo-dynamics. This paper presents a portable multi-channel wireless NIRS device for real-time monitoring of muscle activity. The NIRS sensor is designed miniaturized and modularized, to make multi-site monitoring convenient. ⋯ Besides, the system is designed with high sampling rate so as to monitor rapid oxygenation changes during muscle activities. Dark noise and long-term drift tests have been carried out, and the result indicates the device has a good performance of accuracy and stability. In vivo experiments including arterial occlusion and isometric voluntary forearm muscle contraction were performed, demonstrating the system has the ability to monitor muscle oxygenation parameters effectively even in exercise.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2014
Analysis of EEG to quantify depth of anesthesia using Hidden Markov Model.
Real-time quantification of the patient's consciousness level during anesthesia is an important issue to avoid intraoperative awareness and post-operative side effects. A depth-of-anesthesia (DoA) monitoring method called Bispectral Index (BIS) is generally used for this purpose. However, BIS is known to be inaccurate at the transitory state, and also shows a critical time delay in quantifying the patient's consciousness level. ⋯ Since the evaluation of DoA using HMM is training based method, it have better performance with more training process. Experiments show that HDoA has a high correlation with BIS at a steady state, and outperforms BIS in two ways: (1) shorter delay time in transition state, and (2) higher Fisher Score. The validity of HDoA has been tested by 8 real clinical data.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2014
Classification of serous ovarian tumors based on microarray data using multicategory support vector machines.
Ovarian cancer, the most fatal of reproductive cancers, is the fifth leading cause of death in women in the United States. Serous borderline ovarian tumors (SBOTs) are considered to be earlier or less malignant forms of serous ovarian carcinomas (SOCs). SBOTs are asymptomatic and progression to advanced stages is common. ⋯ Application of the optimal model of support vector machines one-versus-rest with signal-to-noise as a feature selection method gave an accuracy of 97.3%, relative classifier information of 0.916, and a kappa index of 0.941. In addition, 5 features, including the expression of putative biomarkers SNTN and AOX1, were selected to differentiate between normal, SBOT, and SOC groups. An accurate diagnosis of ovarian tumor subclasses by application of multicategory machine learning would be cost-effective and simple to perform, and would ensure more effective subclass-targeted therapy.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2014
Optimising the Windkessel model for cardiac output monitoring during changes in vascular tone.
Algorithms for estimating cardiac output (CO) from the arterial blood pressure wave have been observed to be inaccurate during changes in vascular tone. Many such algorithms are based on the Windkessel model of the circulation. We investigated the optimal analytical approaches and assumptions that make up each algorithm during changes in vascular tone. ⋯ They produced a percentage error of ±31% by maintaining the compliance and outflow terms in the Windkessel model. For any algorithm, the following assumptions gave highest accuracy: (i) outflow pressure into the microcirculation is zero; (ii) end of systole is identified using the second derivative of pressure. None of the tested algorithms reached the clinically acceptable accuracy of ±30%.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2014
Prediction of mortality from respiratory distress among long-term mechanically ventilated patients.
With the advent of inexpensive storage, pervasive networking, and wireless devices, it is now possible to store a large proportion of the medical data that is collected in the intensive care unit (ICU). These data sets can be used as valuable resources for developing and validating predictive analytics. In this report, we focus on the problem of prediction of mortality from respiratory distress among long-term mechanically ventilated patients using data from the publicly-available MIMIC-II database. ⋯ We also find that variables related to respiration rate have more predictive accuracy than variables related to oxygenation status. Ultimately, we have developed a model which predicts mortality from respiratory distress in the ICU with a cross-validated area-under-the-curve (AUC) of approximately 0.74. Four methodologies are utilized for model dimensionality-reduction: univariate logistic regression, multivariate logistic regression, decision trees, and penalized logistic regression.