Rice University logoGeorge R. Brown School of Engineering
 
Electrical and Computer Engineering
 
 

2013 ECE Affiliates Meeting - Poster/Demo Session - Undergraduate Research & Senior Design

 

In Progress 

 

Graduate Research &
Education Innovation
 
 Undergraduate Research &
Senior Designs
 
Startups by Rice Alums and Faculty 

 


Wednesday, April 3, 2013
1:00pm - Martel Hall, Duncan Hall

  1. Integrating Low-Power Sensors with Passive RFID Tags for International Space Station CO2 Monitoring  (Senior Design) - Demo & Poster
  2. Scalable Wireless Alert Generator  (Senior Design) - Demo & Poster
  3. mobileSpiro: A Portable System for Accurate, User-Friendly Spirometry  (Undergraduate Research) - Demo & Poster
  4. An Optogenetic Tool for Automated Neural Circuit Analysis (Senior Design) - Poster
  5. Single Shot Time-Domain Terahertz Spectroscopy (Senior Design) - Poster
  6. Low-cost Vital Signs Monitoring for Neonates in the Developing World (Senior Design) (pdf) - Demo & Poster
  7. Electric Owl II (Senior Design) - Demo & Poster
  8. Neuron signal acquisition (Senior Design) - Poster
  9. An Obstacle-Avoidance System for Power Wheelchairs (Senior Design) - Poster
  10. SMART Belt: A Multi-sensor Seizure Detection Device for Nighttime Outpatient Monitoring (Senior Design) (pdf) - Demo & Poster
  11. Development and Testing of a Novel Myocardial Perfusion Device (Senior Design) - Demo & Poster
  12. Heart Rate Monitoring in Resource Limited Hospitals (Senior Design) - Demo & Poster
  13. RFID Race Timing for Beer Bike (Senior Design) - Poster
  14. Low Energy Fast SIFT Detector Using Mobile GPU (Undergraduate Research) - Demo & Poster


Abstracts
  1. Integrating Low-Power Sensors with Passive RFID Tags for International Space Station CO2 Monitoring
    Authors: Nonso Anyigbo, Daryl Arredondo, Jiwon Choe, Elaine Chung, Kai He, Dr. Joseph Cavallaro, Dr. Gary Woods
    Advisors: Dr. Joseph Cavallaro and Dr. Gary Woods

    Abstract: The microgravity environment on the International Space Station (ISS) causes gases to diffuse irregularly, leading to the formation of hazardous carbon dioxide pockets. To address this issue, Hermes has designed a low-power CO2 sensor system that uses passive radio-frequency identification (RFID) communication for data transmission. The CO2 sensor is controlled by an application microcontroller which provides a tunable data acquisition rate. This microcontroller sends the CO2 sensor data to an RFID microcontroller via a SPI interface, where the data is stored in flash memory. When an RFID reader interrogates the RFID tag, the tag will transmit its stored data to the reader. This system will allow NASA scientists to study how CO2 aggregates in the microgravity environment aboard the International Space Station (ISS), positioning them to address CO2-related health risks which may arise during long-duration space travel.

     
  2. Scalable Wireless Alert Generator
    Authors: Matthew Johnson, Yuqiang Mu, Chris Metzler, Kiran Pathakota
    Advisors: Dr. Ashutosh Sabharwal, Dr. Gary Woods

    Abstract: Malawi is an underdeveloped country with a very high infant mortality rate and grossly understaffed hospitals.  The goal of our project is to create a central display hub that monitors multiple patients of a neonatal ward simultaneously and presents their vitals in an easily interpreted form.  A central display would allow one nurse to monitor many patients at once.  Currently the nurse is forced to go bedside to monitor the babies.   A device that monitors all the patients at once would allow the nurse to spend his or her time only on the patients in need of urgent care, thus greatly improving the hospitals’ ability to treat patients. 

     
  3. mobileSpiro: A Portable System for Accurate, User-Friendly Spirometry
    Authors: Hasitha Dharmasiri, Peter Chang, Gaurav Patel, Ashutosh Sabharwal
    Advisors: Dr. Ashutosh Sabharwal

    Abstract: Lung diseases are among the world’s most critical health issues. Specifically, chronic lung diseases like asthma and chronic obstructive pulmonary disease (COPD) have become major concerns in modern healthcare, and the need for detection and management of these diseases will only continue to grow. We have developed mobileSpiro, a portable, low-cost spirometer intended for patient self-monitoring. mobileSpiro innovates on previous spirometer designs by utilizing the processing capabilities and Internet connectivity of smartphones.  This enables mobileSpiro to provide instant feedback to patients and to allow doctors to access and analyze medical data remotely. mobileSpiro is intended to be released as an end user device and therefore will perform spirometry measurements with sufficient accuracy to receive ISO and ATS certification.

     
  4. An Optogenetic Tool for Automated Neural Circuit Analysis
    Authors: CJ Williams, Storm Slivkoff, Que Kim, Carolyn Ma, Duffy Elmer
    Advisors: Dr. Jacob Robinson

    Abstract: We at Team Illuminate from Rice University are designing and constructing an optogenetic microscopy tool for stimulation, recording, and modeling of in vitro neural cultures. This system will help automate experiments that require parallel neural control and real-time circuit analysis. Our design is motivated by a number of significant shortcomings in current neuroscience research. Primarily, we wish to address the current need for wetlab neuroscientists to rely on outside research groups for data analysis, a collaborative paradigm which slows experimental progress. A comprehensive research tool that can simultaneously stimulate multiple neurons in parallel and automate basic neural circuit analysis will enable wetlab neuroscientists to perform rapid, in-lab analysis and thus accelerate neural circuit research. In collaboration with Dr. Jacob Robinson’s nanotechnology and neural circuit research laboratory, we have devised a computer-controlled, three-component design that addresses the need for simultaneous neural circuit stimulation, recording, and analysis (see figure 1). Stimulation will be achieved by a Digital Micromirror Device (DMD) projector with high enough resolution to stimulate at the single cell level. Patch-clamping equipment will be used to record and measure the resulting voltage spikes from neurons. A user-friendly MATLAB GUI will facilitate the process of modeling neural networks using data from the patch clamp and camera. The scope of our project is to create a functional skeleton of the device by incorporating the projector, camera and microscope optics into a single system that can be further improved by Dr. Robinson’s lab in the future."

     
  5. Single Shot Time-Domain Terahertz Spectroscopy
    Authors: Ayana Andalcio, Patrick Breen, Lisa Anne Hendricks, Tapash Sarkar, Junichiro Kono, Gary Woods
    Advisors: Junichiro Kono and Gary Woods

    Abstract: Team Magneto-Optics is building a single-shot time domain spectroscopy system to measure the spectrum of graphene in high magnetic fields to understand Landau quantization. This quantization of electron energies in a magnetic field parallels the quantization of electron energy in the electric field of the atomic nucleus. Studying Landau quantization requires strong fields (~10-30 Tesla) to induce Landau splitting and Terahertz frequencies to see the energy gaps. Strong magnets are either large and housed in national facilities or are single use and self-destruct. We instead use a unique pulsed magnet (~10 minute duty cycle,10 ms pulse) with a clear optical path and a maximum magnetic field strength of 30T. We use a standard time domain spectroscopy (TDS) scheme. However, the number of time samples needed to generate the desired spectrum coupled with the still relatively low duty cycle of our magnet would require ~400 hours of labor with a standard TDS scheme. By integrating an echelle which breaks a single laser pulse into a series of time-delayed beamlets, we are able to capture a full spectrum in a single shot. THz radiation is generated through optical rectification in a ZnTe crystal, and THz radiation is detected by electro-optic sampling.

     
  6. Low-cost Vital Signs Monitoring for Neonates in the Developing World
    Authors: Nathan Lo, Abhijit Navlekar, Eric Palmgren, Fabio Ussher, Rahul Rekhi
    Advisors: Gary Woods, Maria Oden, Ashutosh Sabharwal

    Abstract: Many hospitals and clinics in the developing world are under-staffed and under-resourced. For instance, a neonatal ward might have 50 patients and only one nurse. Monitoring equipment may be non-existent, and electricity supplies are unreliable. This situation leads to numerous problems, such as: (1) the inability of medical staff to determine when patients are in distress and (2) the lack of a reliable way to record and track trends in vital signs. Several doctors in developing-world hospitals have stated that they would like to have a simple monitoring device that could be attached to patients. The device would report vital signs to a central monitoring station, and/or provide a bedside alarm. In the words of one doctor in a district hospital in Malawi, such a system would “revolutionize” the quality of care that could be delivered.
    Team BioLink seeks to alleviate these issues associated with low nurse-to-patient ratios in under-resourced neonatal hospitals, where the medical staff is incapable of tracking--and responding to--all of the patients' vital signs. We are particularly drive to address the lack of respiratory rate monitoring: a clinical deficit that leads to the deaths of over 14,000 neonates every day in the developing world. A reliable, low-cost, low-power respiratory sensor device will significantly reduce the amount of time that the nurses much spend with each individual patient and will allow much faster responses to medical emergencies. This device will wirelessly interface with a scalable central “brick” to permit round-the-clock tracking and alerting of infants’ health in real-time.
    Thus, we have designed the iNurse™, a low-cost, power-efficient neonatal vitals monitoring system. This device will allow for high-fidelity tracking of infant health in the developing world, allowing for the prevention and mitigation for a variety of clinical disorders and pathologies. Specifically, the needs for this device are as follows:
      • Vitals monitoring key to ensuring health of neonates
      • Low nurse-to-patient ratios preclude proper vitals management
      • Traditional vitals biosensors are too costly to adopt

    We have constructed a preliminary prototype for the iNurse, which has been tested on sample data from adult respiration. We are currently refining the requisite processing algorithms to provide for an accurate alerting method. Stress-testing and motion sensitivity analysis will be performed to assess the system’s robustness in an unstable, crowded, hot, and/or noisy clinical setting. The device will be ultimately tested in the clinic in the Spring of 2013.

     
  7. Electric Owl II
    Authors: Asher McGuffin, Pedro Chacon, Justin Ng, Samuel Whisler
    Advisor: Gary Woods

    Abstract: This project aims to improve the ability of the Electric Owl Unmanned Aerial Vehicle (UAV) to capture data and operate without direct human interaction. Specifically, this year we plan to fully integrate an optical sensor into the UAV control system, so that the information is available to the flight computer, and implement on-board image processing using a BeagleBoard-xM. The integration of an optical sensor is a valuable addition to the Electric Owl system because it adds a means of gathering useful data, which can be used to make better decisions, such as object detection and location. Using this data, the UAV will be able to make better flight decisions, take better pictures, and depend less on human controllers, maps, and GPS. Further, we hope that future design teams will make use of the optical systems that we design and the information produced by our image processing system to make further improvements to the Electric Owl in the future.

     
  8. Neuron signal acquisition
    Authors: Kevin Chu, Xizheng Ma, Di Meng, Xiran Wang, Siyu Wu and Caleb Kemere, Ray Simar, and Gary Woods
    Advisors: Caleb Kemere, Ray Simar, and Gary Woods

    Abstract: This is a multi-channel neuron signal acquisition system that samples at 25 KHz. Signals will be digitally filtered and then displayed on PC.

     
  9. An Obstacle-Avoidance System for Power Wheelchairs
    Authors: Paul Ashla, Ryan Christiansen, Alejandra Galvan, Isabella Gonzalez, Kristen Graber, Rachel Green
    Advisor: Gary Woods

    Abstract: Team Hot Wheels aims to design a patient-driven obstacle-avoiding system for power wheelchairs that automatically detects and avoids collisions between the chair and obstacles in its path. This system is primarily aimed at patients with neurological disorders or loss of visual acuity. Our stereoscopic vision system is currently being improved, and we have firmly established proof of concept with static images. Currently, the challenge is improving real-time operation of the system and implementing the underlying code, which will interpret the visual data and identify an oncoming hazard. We are also programming a microcontroller to process input from an array of ultrasonic range sensors. The ultrasonic microcontroller and the laptop for the stereoscopic cameras will decide whether any obstacles pose an accident risk to the user and will produce warning and stopping signals as appropriate. The most important recent development is the interrupter system, which uses relays to sever the connection from the battery to the motors while activating the parking brake. This operation will be carried out when either of the sensor systems identify an immediate danger, stopping the wheelchair short of the obstacle. In addition, the user interface for the system has been mapped out; this interface includes an LED to indicate a warning, an LED and buzzer to indicate when the system is stopping the wheelchair, and a button to be used for manual override.

     
  10. SMART Belt: A Multi-sensor Seizure Detection Device for Nighttime Outpatient Monitoring
    Authors: Ethan Leng, Mihir Mongia, Charles Park, Tiffany Varughese, Andrew Wu
    Advisors: Maria Oden and Gary Woods

    Abstract: Epilepsy is a chronic neurological disorder characterized by recurrent seizures. The condition affects an estimated 2.2 million people in the United States alone. While several treatment options exist, there is no guaranteed cure. For patients whose seizures are intractable, accurate detection of seizures in an outpatient setting is needed because seizures can present with serious, potentially life-threatening symptoms even if they are not convulsive. Therefore, in the event of a seizure, caregivers need to be alerted quickly so that they may take the appropriate actions. However, existing devices are either designed for clinical use only or have poor performance and excessively high costs. To fill the need for seizure detection that can function in an outpatient setting, we, Team Seize and Assist, have developed the Seizure Monitoring and Response Transducing (SMART) Belt, a novel, wearable, multi-sensor device designed to provide reliable, continuous, night-time monitoring for epileptics. The device is composed of a respiration sensor and an electrodermal activity (EDA) sensor, which are integrated into a single belt that is worn around the torso at the broadest part of the ribcage. An attached electronic module analyzes respiration and EDA data and looks for roughly synchronized changes in both that are indicative of seizure activity, and when a seizure is detected, the SMART Belt wirelessly alerts caregivers via Bluetooth. Although many types of respiration and EDA sensors already exist, to our knowledge this is the first time that the two have been combined into a single device for the purpose of seizure monitoring.

     
  11. Development and Testing of a Novel Myocardial Perfusion Device
    Authors: Ana Estrada, Aditya Kumar, Ian Akash Morrison, Aniruddha Sen, Maria Oden, Gary Woods, Renata Ramos, and Mehdi Razavi
    Advisors: Maria Oden, Gary Woods, Renata Ramos, and Mehdi Razavi

    Abstract: Coronary artery disease (CAD), currently the leading cause of death in the industrialized world, is characterized by occlusion of blood vessels in the heart that prevent proper blood flow and oxygen delivery to the cardiac muscle. End-stage CAD involves the occlusion of numerous capillaries in the coronary vasculature, leading to decreased myocardial perfusion. Current treatment options are limited by the time the disease has progressed to this stage as conventional CAD solutions focus on improving blood flow at the arterial level as opposed to in small arteries, arterioles and capillaries. Given the prevalence and morbidity of end-stage CAD, there is a large need for the development of a viable treatment. Clinical studies have shown that increases in coronary sinus pressure have a direct effect on perfusion pressure. The introduction of a device implanted in the venous end of the coronary vasculature that could increase perfusion pressure could potentially improve blood flow through the coronary capillaries, exposing blocked vessels to blood flow. We are proposing a device-based solution that increases blood flow through the myocardial capillaries through pressure modulation for the treatment of end-stage CAD. Specifically, we are developing a catheter-deployed, axial-flow pump to increase coronary perfusion pressure and retrograde flow into the coronary vasculature. Our initial prototype includes a preliminary model of our device as well as a closed flow loop that mimics the myocardial circulation. Utilizing pressure transducers, flow sensors, and a novel method to mimic perfusion, we have developed a bench-top system to evaluate our device efficacy.

     
  12. Heart Rate Monitoring in Resource Limited Hospitals
    Authors: Gbenga Badipe, Adrian Galindo, Alison Hightman, John Slack, James Kerwin
    Advisors: Dr. Gary Woods, Dr. Maria Oden, Dr. Ashutosh Sabharwal

    Abstract: Africa bears a great percentage of the world’s disease burden but has the least amount of health care workers and funding. With only an average of 11 nurses and 2 doctors per 10,000 people and a neonatal mortality rate more than double that of the US, African countries are facing a health crisis. Besides a lack of health care workers (only 1 nurse for 30 infants in a ward), African countries also suffer a severe lack of critical care medical equipment. Medical equipment such as vital sign monitoring are critical in providing lifesaving care. In fact, research has shown that the mortality rate can be significantly reduced by introducing vital sign monitoring equipment into a hospital setting. More specifically, after further field research, we believe that heart rate monitoring will provide hospital staff the most holistic view of a patient while also enabling them to provide critical care to a patient in distress.
    Periodic data measurement and recording of a patient's heart rate will enable doctors in the developing world to provide more evidence based care by allowing them to more easily determine the effectiveness of their treatments. Our heart rate sensor, when combined with the overall neonatal monitoring system project, will provide a low cost, low power, and reliable heart rate monitoring solution useful for long term monitoring in the developing world.
    We have designed an attachment belt for gathering raw EKG signal, designed an analog front end for filtering and amplifying the raw signal, and developed robust heart rate algorithms for processing this data. We have implemented our solution on a microcontroller. Additionally, we have worked with team SWAG OWLS to define and implement an information passing protocol to send our heart rate data to the larger SWAG system.
    The overall SWAG system exposes our sensor to medical staff in a useful way. It allows for monitoring of multiple babies simultaneously through wireless communication and a tablet interface. It provides immediate alarms in addition to flexible vital sign reporting and trend data. Overall, our device and the larger SWAG system will enable hospital staff to make better decisions for the treatment plans of individual infants, improve neonatal care, and lower the neonatal mortality rate. 

     
  13. RFID Race Timing for Beer Bike
    Authors: Ryan Koehn, Blake Miller, Paul Melvin, Shaun Haby
    Advisor: Gary Woods

    Abstract: Our RFID timing system is composed of four basic components: RFID tags, RFID antennas, an RFID reader, and a computer. During the Beer Bike race, all bikers will be required to wear RFID tags. These tags will be detected by RFID antennas located along the track. All RFID antennas will be connected to a single RFID reader. It is the RFID reader’s job to upload the antennas’ tag read data to the computer, by means of a wired connection. The computer will then process the tag data in order to determine the race results.

     
  14. Low Energy Fast SIFT Detector Using Mobile GPU
    Authors: Blaine Rister, Guohui Wang, Michael Wu and Joseph R. Cavallaro
    Advisor: Joseph Cavallaro

    Abstract: Emerging mobile applications, such as augmented reality, demand robust feature detection at high frame rates. We present an implementation of the popular Scale-Invariant Feature Transform (SIFT) feature detection algorithm that incorporates the powerful graphics processing unit (GPU) in mobile devices. Where the usual GPU methods are inefficient on mobile hardware, we propose a heterogeneous dataflow scheme. By methodically partitioning the computation, compressing the data for memory transfers, and taking into account the unique challenges that arise out of the mobile GPU, we achieve near-realtime detection without compromising the original algorithm. Additionally, we reduce energy consumption by 87%.