Page 68 - ASME DSCC 2015 Program

Technical Program

A Control-oriented Model of Blood Volume Response to Hemorrhage objective Quantitative Assessment of Movement Disorders Through
and fluid Resuscitation Analysis of Static Handwritten Characters
Invited session paper. DSCC2015-9847 Invited session paper. DSCC2015-9974
Ramin Bighamian, Jin-oh Hahn, University of Maryland, College Park, MD, Naiqian Zhi, Beverly jaeger, andrew gouldstone, rifat Sipahi,
United States, Andrew T. Reisner, Massachusetts General Hospital, Boston, Northeastern University, Boston, MA, United States, Samuel frank, Boston
MA, United States University Medical Center, Boston, MA, United States
This paper presents a control-oriented model of blood volume response Movement disorders associated with Essential Tremor (ET) and Parkinson’s
to hemorrhage and fluid resuscitation that can be potentially utilized in disease (PD) can negatively impact use of the upper limb for many precision
closed-loop control of fluid resuscitation. A unique characteristic of the pro- tasks, including handwriting. Both ET and PD can be assessed through
posed model is that it is built to ensure structural parsimony while retaining clinical tests which are, however, relatively subjective. This assessment ap-
physiological transparency. To accomplish this characteristic, blood volume proach possesses inherent logistical and resolution limitations. To address
regulation in the body to external perturbations of hemorrhage and fluid this, here we present objective computerized metrics intended to assess
resuscitation was modeled as a low-order control system in which the fluid and quantify the extent to which static writing samples display the effects of
transfer between blood and interstitial fluid is governed by a proportion- ET and PD. Specifically, these metrics are tested in their ability to mea-
al-integral controller. This in essence resulted in a minimal model with four sure tremor by comparing unaffected writing samples with those affected
parameters to be adapted to each individual. The validity of the proposed by artificially induced tremor on healthy subjects, and also by comparing
model was tested using data available in the literature. The results indicated healthy writing samples with symptomatic writing samples collected from
that the proposed model was able to reproduce the blood volume response PD patients reporting micrographia. Our findings indicate that the present-
to hemorrhage and fluid resuscitation with high fidelity: on the average, the ed metrics can be utilized for assessment, leading to a toolset capable of
prediction error was only 1.53 ± 11.5 %, thus strongly supporting our claim that objectively monitoring static handwriting changes associated with symptom
it can be used as viable basis for the design of closed-loop fluid resuscita- variations in ET and/or PD patients.
tion controllers.
ConTRIBuTED SESSIon
A Real-Time Brainwave Based neuro-feedback System for Cognitive 1-15-1 fA3 Adaptive and Intelligent Systems Control
Enhancement George Bellows E 10:00am–12:00pm
Invited session paper. DSCC2015-9855
reza abiri, joseph McBride, xiaopeng Zhao, University of Tennessee, Session Chair: Marcello Canova, Ohio State University
Knoxville, TN, United States, Yang Jiang, University of Kentucky, Lexington, Session Co-Chair: Benjamin Gruenwald, Missouri University of Science
KY, United States and Technology
Brain Computer Interface (BCI) provides a pathway to connect the brain to
A non-Conservative Approach for the Estimation of the Region of
external devices. Neuro-rehabilitation provides advanced means to assist
operation of uncertain Adaptive Control Systems
people with movement disorders such as post-stroke patients and those
Contributed regular paper. DSCC2015-9665
with lost limbs. While much progress has been made in neuro-rehabilita-
tion as assistive devices, few studies had examined mental rehabilitation Mario luca fravolini, Antonio Moschitta, University of Perugia, Perugia,
assisted by BCI such as memory training using neuro-enhancement. It Italy, Tansel Yucelen, Benjamin Gruenwald, Missouri University of Science
and Technology, Rolla, MO, United States
should be noted that many patients with physical disabilities also suffer
cognitive difficulties. On the other hand, cognitive decline can also be the A challenging problem for Model Reference Adaptive Control Systems is the
result of normal aging without brain injury nor diseases. Here, we propose accurate characterization of the transient response in the presence of large
a novel real-time brainwave BCI platform for enhancing human cognitive by uncertainties. Early prior research by the authors has demonstrated that
designing and employing a personalized neuro-feedback robot. Short-term using a projection mechanism for parameters adaptation, the tracking error
memory and attention are among the most important cognitive abilities dynamics behaves as a linear system perturbed by bounded uncertainties.
which manifest in many mental diseases. A social robot is integrated into This has the advantage that the stability analysis can be cast in terms of a
the BCI system to provide feedback based on individual’s brainwaves and convex optimization problem with LMI constraints so that efficient numerical
memory performance. As a simple scenario of memory task, real-time EEG tools can be used for the adaptive controller design. A possible limitation of
signals will be monitored during a visual object memory task. Our novel the approach is that the design is restricted to quadratic control Lyapunov
neuro-feedback system has great potential as a neuro-enhancing device for functions that could produce conservative results in the estimation of robust
cognitive rehabilitation. invariant regions for the adaptive system operation. In this paper this ap-
proach is extended to arbitrary high degree polynomial Lyapunov functions
by translating the design and performance requirements in terms of Sum of
Square inequalities and then using SOS optimization tools for the design.
In this effort the new SOS approach is introduced and compared with the
previous one. A numerical example based on the short period longitudinal
dynamics of the F16 aircraft is used to demonstrate the efficacy of the novel
method.
68