For Prospective Students:

  Apply Online
  Faculty and Areas of Research
  Sample Projects
  Program Alumni and Past Projects
  SUNFEST in the News

For Accepted Students:

  Required Workshops
  Organizational Matters
  Current Students & Projects
  Student FAQs
  Student Resources
  Library Resources

Contact SUNFEST:

Mailing Address:
203 Moore Building
University of Pennsylvania
200 S. 33rd Street
Philadelphia, PA 19104-6314

Find us on Facebook!


Summer 2013

Alumni 2013

George Major Chen (Biomedical Engineering) – Johns Hopkins University

Model-Based Conformance Testing for Implantable Pacemakers

Advisor: Dr. Rahul Mangharam

Abstract: Between 1990 and 2000, over 600,000 implantable cardiac pacemakers and cardioverter defibrillators were recalled. 41% of these devices were recalled due to device software issues. Software-related recalls are increasing with the growing complexity of medical device software, which is responsible for the life-critical operation with the organ. Currently, there are no formal methods to test and verify the safety of implantable cardiac device software. To this effect, a pacemaker-testing platform has been developed to automatically verify if the software in a pacemaker is functioning appropriately and determine if the pacemaker implementation conforms to the device software design specifications. A testing methodology was developed where tests were automatically generated from a model of the pacemaker that satisfied the specifications. These tests checked the software implemented in the physical pacemaker were in conformance with the design specifications and ensured safe operation. This paper outlines the steps used to create this testing platform, as well as the steps used to construct a pacemaker model for testing. By using this test framework as a standard for medical device testing, the US Food and Drug Administration (FDA) will potentially have a more streamlined method to certify the safety of medical device software. Project report | Presentation slides

Cedric Destin (Electrical Engineering) – Temple University

Altitude Estimation for Micro Aerial Vehicle in a Complex Environment

Advisor: Dr. Mark Yim

Abstract: Development done in Mechatronics has allowed Aerial Vehicles (AVs) to be scaled down and be used for indoor and outdoor environments. Micro Aerial Vehicles (MAV) provide small size and great flying ability allowing professionals to use them for a wide range of applications. Flying Micro Aerial Vehicles in complex environments such as indoor settings requires a semi-autonomous system able adjust to surrounding objects. Therefore to address the issue a network of sensors is implemented to monitor and control the altitude of the system regardless of the environment.

Our objective for the summer is to assess various types of sensors and observe some of the constraints that we must consider when monitoring altitude for a flying vehicle. The sensors evaluated were an IR sensor from Sharp Electronics, a Sonar sensor LV-MaxSonar-EZ1; both sensors are range finders typically used to detect objects; and finally a barometric sensor. The MPL3115A2 by Freescale Semiconductor was used to analyze the effects of the down-wash of the propellers on a barometer. The research done through the summer helped analyze the accuracy that is obtained when using each sensor. The data obtained helped quantify the maximum range, data rate, and type of noise for each sensor.

In conclusion, recent research showed that Micro Aerial Vehicles have great potential in the future; by implementing the appropriate sensors, we can optimize our system for more applications. Finally, an experiment is performed using the laser range finder to test whether the information gained from the characterization of the sensors can produce greater accuracy when exploring a new area with the quadrotor. Project report | Presentation slides

Rodolfo Finocchi (Biomedical Engineering) – Johns Hopkins University

Development of a Post-Traumatic Osteoarthritis Model: Studies on the Mechanical and Biochemical Effects of Injury on Cartilage Tissue Analogs

Advisors: Dr. Robert Mauck and Dr. George Dodge

Abstract: Articular cartilage can be damaged by traumatic injury, is slow to grow and repair after injury, and can eventually be thinned or completely worn out, resulting in debilitating pain and reduced joint motion. This condition, called post-traumatic osteoarthritis (PTOA), is highly prevalent and affects approximately 6 million individuals with both physical and economic consequences affecting the well-being of the patient. Various in vitro, ex vivo, and in vivo models have been developed to better understand different mechanical and biochemical properties of cartilage affected by PTOA. In this work, we create an in vitro model of PTOA by examining the effects of sudden impact and continuous physiologic loading on the structural and biochemical properties of both native and engineered cartilage. We also evaluate the anti-inflammatory and repair-inducing effects of various chemical compounds (anti-apoptotic, inhibitors of matrix loss) in this engineered cartilage model of impact. Our results showed that a few of these compounds can have positive, therapeutic effects on construct properties after impact or physiological loading. To study these conditions further, combinatorial studies involving the use of both immediate injury and continuous loading, as well as those involving the use of multiple compounds are underway. Project report | Presentation slides

Ty’Quish Keyes (Biomedical Engineering) – Morehouse College

Mesenchymal Stem Cell Response to Static Stretch on Poly-L-Lactide Scaffold

Advisors: Dr. Robert Mauck, PhD

Abstract: Mesenchymal stem cells (MSCs) are multipotent stem cells that have been considered for an increasing list of therapeutic practices, not simply because of their inherent ability to differentiate into connective tissues including bone, fat and cartilage, but furthermore due to their trophic and anti-inflammatory effects which contribute to healing and tissue regeneration. MSCs are often affected by the growth factors they encounter as well as the physical cues from their cellular microenvironment. These microenvironmental cues are important for tissue engineering, where stem cells must be able to differentiate down specific lineages and organize into tissue like structures. One way of investigating the effects of MSCs’ microenvironment is through seeding of polymer nanofibrous scaffolds. These scaffolds can be generated with a highly aligned nanofibrous structure that mimics the native microenvironment of tendons, ligaments and fibrocartilages. Further, these scaffolds can be engineered to have a crimped or wavy structure, which is a property of native tissue microarchitecture that is known to be important for the bulk tissue mechanical response and is likely a regulator of cellular mechanotransduction. In this study, we examine the effects of static stretch on MSCs using aligned or crimped electrospun polymer scaffolds. We also examine the role inhibitors or activators of cellular contractility play in regulating nuclear deformation on MSC seeded scaffolds. Here, we report that the Poly-L-Lactide(PLLA) crimped scaffold is capable of reducing nuclear deformation under static stretching. We also report that increasing contractility, with Lysophosphatidic acid (LPA), or inhibiting the contractility and the actin cytoskeleton, with Y27632 or Cytochalasin D, further prevents nuclear deformation. Due to the low yield of PLLA scaffold dynamic loading of MSCs has not yet been explored. Future studies must be conducted in order to evaluate MSC response at higher static strain as well as MSC response to dynamic loading on PLLA scaffolds. Project report | Presentation slides

Andrew Knight (Physics) – Norfolk State University

The Investigation of Iron Oxide Nanoparticles as a Novelty for Smart Windows

Advisor: Dr. Shu Yang

Abstract: Smart windows refer to electronically switchable glass which change how light is transmitted when a voltage is applied. With steadily rising energy costs in the United States smart windows are an excellent choice to reduce energy consumption.

One smart window technology, suspended particle devices (SPDs) via Fe3O4 (iron oxide) nanoparticles, has presented itself as a viable option for future window applications. However, there are several issues that must be resolved before this technology can be fully implemented. That is Fe3O4 nanoparticles have a very low transmittance (<40%), particle dispersion is not uniform, and the difficult particle aggregation after a voltage is applied. In this study, we further investigate Fe3O4 nanoparticles and a Fe304@SiO2 (Silica) core-shell nanoparticle. Here, we report that by decreasing the concentration of Fe3O4, the transmittance increases. We were able to achieve better particle dispersion and aggregation by reducing the amount of hydrogen peroxide used to create the Fe3O4 nanoparticles from 2.0ml to 1.5ml. In addition, through SEM (scanning electron microscope) image, we successfully created Fe304@SiO2 core-shell nanoparticles. Our next step involves testing its reliability. Project report | Presentation slides

Quentin Morales-Perryman (Chemical Engineering) – Hampton University

Autonomous Robots

Advisor: Dr. Daniel D. Lee

Abstract: In 2010, Australia hosted an event called the Multi Autonomous Ground-robotic Competition (MAGIC 2010). The goal of MAGIC 2010 was for teams of students to create a group of autonomous robots with the ability to explore terrain and execute given missions such as surveying and mapping an unknown environment. The goal of our project was to build on the foundation developed by MAGIC 2010, which simply opened the door to the possibilities and difficulties that existed in autonomous robotics. To do the project, we needed a way for a group of robots to map unknown terrain without the use external mapping tools, such as GPS. We also needed a way for the robots to coordinate and communicate information with each other so they could work together autonomously. To meet the goal, we made a team consisting of ground robots paired with quadrotors. The ground units are the processing station where the majority of the computation is done. The quadrotor flies implements a tag identification system in order to locate the position of any surrounding ground robots and sends the information to the paired ground robot for processing. After further testing and improvement, we believe that our work will be used to aid in the field of search-and-rescue and assist in the exploration of dangerous environments. Project report | Presentation slides

Samantha Muñoz (Biomedical Engineering) – Vanderbilt University

Time-Varying Network Model of Neurodegenerative Disease Spread in Biological Neural Networks

Advisor: George J. Pappas

Abstract: Although some cognitive decline over time is considered normal, neurodegenerative diseases such as dementia can exacerbate loss of memory and physical functionality, and even lead to death, especially within the elderly population. Currently, treatment for these types of diseases is limited by inadequate understanding of how these diseases progress throughout the brain. Emerging studies show that the spreading and building up of beta-amyloid proteins in certain grey matter regions of the brain could be a possible indicator of disease progression, specifically in Alzheimer’s. In order to assist with potential treatment advances, we created a dynamic spreading model that will better predict the progression of Alzheimer’s disease by looking at covariance of grey matter densities in different regions of interest within the brain of patients with varying stages of dementia The hope is that this model is better able to predict specific areas of the brain where the grey matter densities will change due to the disease process. Future studies into the buildup of proteins that affect grey matter density could use this model to target the specific areas for intervention with new treatments. Our model could also be applied to the spreading processes of other diseases and be used to help patients and families better understand and prepare for the progression of their illness. Project report | Presentation slides

Gedeon K. Nyengele (Electrical Engineering) – Georgia Perimeter College

A Mobile System to Monitor Neonatal Nursing Characteristics (Neonur)

Advisor: Dr. Jay Zemel

Abstract: Neonatal development is considered as a complex process to monitor because, due to the inability of neonates to effectively communicate, the majority of the information about neonatal physiology needs to be extracted by electronic means. Studies have shown that information about infant’s behavioral and physiological states can be acquired by analyzing parameters related to the sucking pressure and its frequency. Multiple attempts have been made in the development of devices capable of monitoring neonatal behaviors such as breathing and feeding. However, the use of those devices is usually limited due their prices, sizes, and ease of use. This paper proposes a design of a convenient, mobile, and energy efficient monitoring system (Neonur) that could be easily assembled and attached to a baby bottle. The monitoring system is equipped with a pyroelectric breathing sensor constructed out of polyvivylidene fluoride films and a standard disposable micro-electro-mechanical pressure sensor widely used medical applications. The pyroelectric breathing sensor provides valuable information about the infant’s respiratory state by generating electric currents that are proportional to the magnitudes of the small changes in temperature on the films due to the infant’s exhalation. Data gathered from the breathing and sucking pressure sensors is saved on the on-chip memory and later transferred to a computer via the Universal Serial Bus (USB). The results indicate that this device is best suited for monitoring neonatal breathing and feeding characteristics, is easy to operate, and is cheap to produce. Project report | Presentation slides

Gabriela Romero (Biomedical Engineering) – Worcester Polytechnic Institute

Nickel Chloride-Mediated Protein Attachment to Molybdenum Disulfide For Biosensing Applications

Advisor: A.T. Charlie Johnson

Abstract: Single- and few-layer Molybdenum disulfide (MoS2) thin films, which have recently been synthesized for the first time, are of great interest for potential applications due to their two dimensional structure and electronic properties. With a bandgap of 1.8 eV, conduction through this material can be tuned between on and off states; a property that graphene, a more studied two dimensional material, does not possess. Furthermore MoS2, presents high thermal and chemical stability, which allows the creation of high-performance nano-electric devices such as field effect transistors (FETs), which could be used as ultrasensitive sensors for clinically-relevant proteins and other biomolecules. These kinds of sensors are currently fabricated with carbon based materials such as nanotubes and graphene and have detection limits at the femto-molar levels. However, the bandgap of MoS2 could allow for lower concentration detections. This study investigates the process of Nickel Chloride (NiCl2) mediated protein attachment to exfoliated MoS2 flakes using different methods of purification and different concentrations of NiCl2.Testing our mechanism is ongoing and will specify the most beneficial conditions of NiCl2 needed to attach the highest density of proteins to the MoS2. This is the first step towards building a biosensor based upon molybdenum disulfide. Project report | Presentation slides

Basheer Subei (BioEngineering) – University of Illinois at Chicago, Cook County

Microcontroller-based General Platform for a Wireless Brain-Computer Interface

Advisor: Dr. Van der Spiegel

Abstract: Recent advances in embedded wireless technology opens the door to fully portable closed loop brain-computer interfaces (BCI) that give neuroscientists the ability to run BCI experiments on primates in an unconfined natural setting. Currently developed portable BCI platforms, however, do not incorporate customizability in their designs, requiring the researcher to modify the device’s firmware whenever they need to tweak the device’s settings. This means that the researcher has to have designed the BCI him/herself or require them to have the BCI designer tweak the settings. Needless to say, this hinders the prospects of BCI platforms being widely deployed and used by neuroscientists for experiments. This paper presents a microcontroller-based BCI design that aims to provide a general wireless BCI platform that incorporates customizability and the ability to tweak settings over-the-air from a simple PC application interface. The limited on-air data rate (2 Mbps) of the wireless transceiver currently limits the data transfer to four 12-bit resolution ADC channels for recording at 20ksps and four DAC channels for stimulation. In-house development of an Ultra-wide Band (UWB) wireless transceiver with a much higher on-air data rate has already started, which would allow for a greater number of recorder channels at a much higher sampling rate. Project report | Presentation slides


See also the alumni and past projects from: 2013 | 2012 | 2011 | 2010 | 2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003 | 2002 | 2001 | 2000 | 1999 | 1998 | 1997 | 1996 | 1995 | 1994 | 1993 to 1986