Texas MRC Research Projects


We propose to employ near-infrared spectroscopy (NIRS) technology to continually map hemorrhagic contusions on the surface of the brain after traumatic brain injury (TBI). This is a novel approach for monitoring the evolution of contusions, while validating results with computed tomography (CT). In contrast to CT that can only be performed infrequently, NIRS will enable the continual bedside monitoring of contusion evolution that is of critical clinical importance for life saving interventions. In parallel with our clinical goal, we propose to build and test a NIRS imaging instrument prototype that is designed to overcome current commercial technology limitations and have the capacity to alert clinical staff in real time when contusions grow rapidly. This technology is foreseen to be of immediate interest to the military, sports medicine personnel and clinicians across all trauma ICU facilities.

Georgios Alexandrakis, Ph.D. UTA
Associate Professor – Bioengineering

Duncan L. MacFarlane, Ph.D. UTD
David C. Smith, M.D. THR

No Comments

We propose to create a miniaturized watch-based energy expenditure monitoring system. Our proposed system will combine heart rate acquired from a wrist-worn photoplethysmogram (PPG) sensor and physical activities using motion sensors in a watch. We will create predictors and fitting functions that use heart rate and activity type/level to estimate the energy expenditure. Our fitting functions customized for individuals will be validated in conjunction with an integrated metabolic measurement system. We will focus on advanced motion artifact rejection techniques that will incorporate the notion of context. Context includes noise models obtained from the motion sensors, as well as the models for the heart rate collected in the past. Knowing that the heart rate cannot fluctuate rapidly, during the periods when the noise due to the motion artifacts increases significantly, we will use the heart rate detected prior to the increased motion-induced artifacts to guide the signal processing modules. We will create signal processing techniques to identify physical activities from the motion sensors. We will validate our energy expenditure monitoring system against an integrated metabolic measurement system at UT-Arlington on (N=20) human subjects.

Roozbeh Jafari, Ph.D. UTD
Associate Professor – Electrical Engineering

Christpher Ray, Ph.D. UTA

No Comments

Device-related infection is a leading cause of failure in many medical devices such as joint prostheses, catheters, prosthetic heart valves and pacemakers. For example, the infection rate of orthopedic implants is 4-5%. Given the aging population and increased need for medical devices, there is an estimate of 100,000 device related infections, which poses a major liability of over $4.0 billion annually on the healthcare system. Infection of medical devices is caused by formation of bacterial biofilm on the device. Biofilms are colonies of bacteria attached to a surface in a self-produced and highly protective matrix. Biofilms on medical devices are the source of major challenges for medical settings. This is because bacteria in biofilms are >1,000 times resistant to antibiotics compared to the free-floating bacteria. Eradication of an infected implant through regular antibiotic therapy is not feasible, and often times the only solution is a revision surgery and second implant, which has a higher chance of infection; and its failure may result in amputation in orthopedic cases. FDA has been concerned about device infection and held a conference as recent as February 2014. Generally, the best solution for fighting biofilms is preventing their formation in early stages, which requires highly sensitive sensors, and is currently lacking. We, a team of PIs with complementary Engineering and Microbiology background, propose to develop and commercialize a marker-free integrated Micro/Nano-Electromechanical (M/NEMS) sensor with fully electronic readout. The sensor will detect biofilm formation in very early stages and actively deliver highly localized doses of antibiotic to eliminate bacteria before biofilm formation. This highly sensitive sensor will find applications in “intelligent implantable devices”, in an over $10 billion market.

Majid Minary, Ph.D., UTD
Assistant Professor-Mechanical Engineering

Siavash Pourkamali, Ph.D. UTD
Woo-Suk Chang, Ph.D. UTA
Jerry Simecka, Ph.D. UNTHSC
Randall Todd Richwine, Ph.D. THR

No Comments

Annually in the US more than 1/3 of adults, ages 65 and older, experience falls. Many of these fallers start using a cane or rolling walker (RW) following a fall. In senior living communities, a RW is the most commonly used ambulatory device. Among the older RW users, pain at the wrists and upper and lower back often occurs. Unfortunately, the rate of falling is still high (nearly 40%) amongst these RW users. The side effects (particularly the risk of falls) of using a RW are greatly associated with incorrect RW height (causing a different grip strength distribution while holding the RW), inappropriate user posture (causing different pressure/strains on the walker/ground), changes in gait patterns (causing a decreased gait speed) and poor maintenance of the RW tips (wheels) (causing different shear force on the ground). However, currently there are no instruments or devices available to enable a clinician to monitor how a RW user uses their RW for daily mobility. Therefore, in this study, a SmartWalker will be designed and instrumented in our engineering laboratory in order to identify the improper RW use before it causes a fall or other fall‐related medical problems. Different sensors will be installed on the SmartWalker. These sensors will monitor how the user holds the handgrips of a walker, how the user’s posture affects ground‐reaction forces during standing or walking, how fast the user moves the walker during ambulation and if and how pressure is distributed on the four tips (wheels) of the walker. Data from the sensors will be acquired by a Data Acquisition Device (DAD) and wirelessly transmitted to a PC for data storage and to monitor the user’s status while using the walker. Further, this study will evaluate the use of the SmartWalker longitudinally at two local senior living communities using common clinical assessment tools such as reliability and validity tests, feasibility analyses and correlation studies. The SmartWalker is expected to be an assistive ambulatory device as well as a dynamic, real‐time, data‐traceable monitor of the older walker user’s grip strength, posture, gait speed and overall maintenance for the prevention of walker‐user related falls and side effects.

Howe Liu, Ph.D. UNTHSC
Associate Professor – Physical Therapy
Haiying Huang, Ph.D. UTA
Arvind Nana, M.D. THR

No Comments

Navigation around an organ during surgery is not an easy task as the visual view is not always reliable. Visual clarity is often limited due to existence of blood and fat tissues. Also, every case and every organ is unique in terms of size, shape, location and so on. In particular, in 10-20% of the open heart surgery cases, surgeons have difficulty locating the exact blood vessel with the blockage.
Our proposal is to design an imaging system that helps surgeons to navigate easier and more accurately around the heart especially during bypass surgeries. Our twin goals will be to:
(a) find/design a small set of markers (e.g. colored and/or RFID-based) that provide a clear indication of their positions, and
(b) use the markers and a hand-held device to produce the 3D image of the organ with accurate tracing of the markers and device.
Achieving these goals will require design of markers, a navigational hand-held device and the use of image processing techniques to overlay a pre-recorded X-ray (or MRI) image with real-time images from a camera during surgery.

Principal Investigator:
Mehrdad Nourani, Ph.D. UTD
Professor & Associate Department Head – Electrical Engineering

Kambiz Alavi, Ph.D.
James B. Park, Ph.D.

No Comments

About 2.7 million individuals are living in the U.S. with Paroxysmal Atrial Fibrillation (AF), and that number would increase to 15.9 million by 2050 with more than 50% of these numbers accounted for by patients who are more than 80 years old. Identifying AF early on and treating them promptly increases significantly the clinical outcomes. Today, most of the cases of AF are identified when patients come to the clinic complaining of palpitation or accidentally when they undergo routine or other heart checkups. AF is a major cause of stroke. Typically, the prevention of strokes in high-risk patients requires permanent anticoagulation, with a clear risk of life­ threatening bleeding. However, if a patient could be identified as going into AF, immediately then they could start taking anticoagulants and since modern anticoagulants have an immediate effect, the stroke can be prevented without the risk of bleeding. Similarly, antiarrhythmics have serious side effects including arrhythmias and sudden death. If one could identify when AF occurs, a “pill in the pocket” approach could be used and the patient only takes the antiarrhythmic when the arrhythmia develops, saving the patient cost and side effects. This scenario clearly warrants for a device that can warn a patient when going into AF to seek immediate attention.

The proposed research is the design and development of an AF warning system. One of the important components of our proposed work is the development of a model for determining whether there is an abnormality in the heartbeats of the person being monitored, so that the system can provide a warning to the person when the abnormality is observed. We propose to design a hardware device integrated with a software algorithm based on machine learning that uses information from the ECG signal captured by the hardware device to detect AF. The proposed device will have direct impact on the population either by saving their life or by enhancing their quality of life.

Principal Investigator:
Lakshman S. Tamil, Ph.D. UTD
Professor – Electrical Engineering

J-C. Chiao, Ph.D. UTA
Benjamin Levine, M.D. THR


No Comments

Our Partners