SURF-IoT 2015 Concludes
August 31, 2015 / By Sharon Henry
SURF-IoT Fellows make final project presentations at Calit2.
Irvine, August 31, 2015—The 2015 Summer SURF-IoT program concluded Thursday when final presentations were delivered. This was the eleventh consecutive year, the 10-week SURF-IT student summer research program has been held at UC Irvine
SURF-IoT – the Summer Undergraduate Research Fellowship in Internet of Things – provided 11 UC Irvine students a hands-on research experience under the mentorship of faculty and graduate students. The program is jointly sponsored by Calit2 and UCI’s Undergraduate Research Opportunities Program.
Full abstracts of this year’s fellows and projects are listed below:
CLOUD-BASED TOOLS TO EMPOWER INTERDISCIPLINARY
Research: A Case Study on the Mesoamerican Color Survey Data Archive
Fellows: Nathanial A. Benjamin, Stephanie M. Chang, Yang Jiao
Mentors: Kimberly A. Jameson, Sergio Gago, Sean Tauber
Our SURF-IoT project aims to develop a rich public-access database as an internet resource to facilitate scholarly research in an interdisciplinary and empirical manner. The database derives from The Robert E. MacLaury Color Categorization Archive – a large corpus of heretofore unpublished cognitive anthropology survey data consisting of irreproducible observations of color categorization behaviors, from a large number of monolingual participants responding in over 116 indigenous languages. This resource will enable researchers to easily query and manipulate the archive’s data to find linkages necessary for original research in fields such as: anthropology, psychology and linguistics. The archive’s information will be accessible to researchers from a wide variety of disciplines, permitting scientists to address questions related to the formation of individual and shared color concepts, including their evolution and dispersion across ethnolinguistic groups. Towards these goals, we explored three aspects of archive development: (1) Optical Character Recognition (OCR) approaches for automatically converting handwritten raw data into computer addressable datafiles, (2) Empirical “crowdsourcing” survey approaches using large informant groups for transcribing handwritten data, and (3) development, from the ground up, of an underlying Content Management System and a front-end website as a user-interface for the database. We summarize advances along these lines of development to illustrate progress achieved towards our goal of creating a free-access research archive of highly valuable data. The archive we are creating promises to be a long-lasting and valuable resource for researchers seeking to study cognition and categorization across ethnolinguistic groups.
SURF-IoT Fellow Jungkyu Park is also a member of this group.
HEALTH360
Fellow: Thomas Wurzer
Mentor: John Billimek, Anmol Rajpurohit
Health360 is a web application that investigates the feasibility of n-of-1 trials, which are trials in which the patient is the main person that is tested with a certain medication. In today’s world, there are many different kinds of medications, but those that need them often do not take them as prescribed. Health360 is a method that can help medication users take their medication according to their specific needs. This will be done by tracking the user’s health while they take the medication. When they visit their doctor, they will view this data that they have collected and use it to determine the best amount of medication to take. We hypothesize that by using this method of health tracking, we will be able to better care for a patient’s health and be able to administer a good amount of medication for them. The results of this study are important because it means that doctors will be able to tell their patients more easily what the most effective medication will be under their specific circumstances. In the future, we plan to test this application with real users and real data to see how well this method works.
M2M-TRACKING: ENCOUNTER-BASED COLLABORATIVE
Localization using M2M-sisted Particle Filter
Fellow: Albert Lau
Mentors: Mohammad Al Faruque, Hsinchung (Andrew) Chen
Global positioning system has been a useful tool to accuracy map a person location. The problem with global positioning system is that it consumes a lot of energy and cannot be use indoor. In this project, we will try to use a Bluetooth low-energy, which is known to last for months on a single coin battery, to predict a person location when he or she is indoor. We will be using three sensors, an accelerometer, a magnetometer, and a gyroscope, to help track an object. Conditional random field or particle filter will be used to help pin point the object location. Once we find if Bluetooth low-energy can be a feasible device for indoor localization, it can be used in multiple applications from something as small as finding a lost item or something as big as building smart homes.
QUALOSCOPY
Fellow: Luke Raus
Mentors: Willam Karnes, Donald J. Patterson
Colorectal cancer is the leading cause of cancer deaths among nonsmoking Americans. A ten-year window of opportunity exists when colorectal cancer can be prevented by finding and removing colorectal polyps. Colonoscopy is associated with a 70-90% reduction in colorectal cancer incidence. Approximately 90% of colorectal cancers occur in people who are not screened, whereas about 9% of colorectal cancers occur in patients who are up-to-date with their colonoscopies. Approximately 80% of these "interval cancers" are believed to be related to failure of colonoscopists to find and completely remove polyps. Quality colonoscopy has thus become a major emphasis of Medicare's new quality measures program. Colonoscopists are now required to report several quality indicators with associated benchmarks that will determine their level of reimbursement. Qualoscopy is a web-based application we are designing to help gastroenterologists painlessly collect, monitor, and report their quality measures. De-identified data will be made available for research purposes.
USING WEARABLE AND CONNECTED SENSING: PLATFORMS TO INFORM PEDIATRIC INFORMATICS
Fellow: Wilmer Domingo
Mentor: Gillian Hayes
In the past few years, activity trackers have been making their debut - sensing platforms equipped to track health indicators, such as movement, heart rate, skin temperature and UV exposure. Such platforms are ever-present, connected health tracking platforms. While current health solutions involving activity trackers concentrate on fitness tracking, activity trackers may also best suit the physician’s need to follow up with their patients after the doctor’s visit. It is impractical for a physician to constantly be at their patient’s side, but the inability to do so leads to issues of adherence to recovery regimens. To address this problem, we explore the application of activity trackers in assisting physicians with such patients. We have developed the BrainTrack platform, which tracks and assists in the treatment of children who have recently suffered from concussions and are suffering from cognitive and physical impairment, a suitable target for activity trackers. To accommodate their situation, the platform includes a mobile application with a clear and simple design, which provides guidelines for patients toward their recovery and a channel through which patients may express concerns or feedback on their progress. Within this framework, we use the Microsoft Band directly message patients and record health data to be sent to physicians, which will improve administration of the concussion recovery protocol. Future work will include working closely with the local children’s hospital and testing our prototype in a clinical setting.
VESTIBULAR REHABILITATION USING WIDE-ANGLED HEAD MOUNTED DISPLAYS WITH STEREOSCOPIC 3D
Fellow: Thomas Lee
Mentors: Hamid Djalilian, Crista Lopes, Marlon Maducdoc
The vestibular system of the inner ear is responsible for sensing orientation and rotation of the head. Additionally, the system primarily drives reflexes to maintain stable vision and posture. A normal vestibular system can adjust reflexes based on varying situations but adaptation to a loss of vestibular function may be slow and result in sensations of dizziness and vertigo. Patients with vestibular dysfunction may also experience visual vertigo (VV) or visually induced dizziness due to the mismatch of perceived visual and vestibular stimuli. The current standard of care for vestibular dysfunction is vestibular rehabilitation exercises to hasten adaptation. VV symptoms are believed to be caused by an excessive reliance on visual cues for perception and postural stability in patients with vestibular dysfunction. Subsequently, studies have shown that VV symptoms can only be reduced if vestibular rehabilitation exercises are combined with immersive moving visuals or optokinetic stimuli. Recent studies demonstrated that graded exposure to optokinetic stimuli causes adaptive changes and decreased reliance on visual cues, thus, improving VV symptom . Another preliminary study also showed that patients with chronic vertigo improved after repeated exposure to optokinetic stimuli. Furthermore, experiments demonstrated that the combination of vestibular exercises and optokinetic stimuli showed greater improvement in VV symptoms and postural stability compared to vestibular rehabilitation exercises alone. One of the main challenges of implementing optikinetic stimuli is the cumbersome requirement of large screens and equipment. Additionally, the stimuli should be a realistic and interactive environment in which the patient is immersed and result in adaptation and desensitization. Head mounted displays have been suggested to provide an immersive optokinetic stimulus that can be easily implemented in clinical settings in conjunction with vestibular rehabilitation exercise programs. The purpose of our research project is to design and develop a realistic and immersive virtual environment and use Oculus Rift VR goggles to deliver graded, standardized optokinetic stimuli to our patients with vestibular dysfunction. The Oculus Rift goggles are a small wide-angled head mounted display (HMD) with stereoscopic 3D. The goggles are also capable of tracking orientation and head movement. They will provide an immersive optokinetic stimulus without the use of large monitors or screens. Oculus Rift VR goggles are typically used for gaming; however, we saw the potential medical application of the device. We plan to directly engage Ocular Rift into our projects as the company is based only a few miles from UC Irvine's campus. The Donald Bren School of Information and Computer Sciences will play a key role in the design and development of a customized virtual environment. To date, no other study has used HMDs with a customized virtual environment in conjunction with vestibular rehabilitation exercises. Additionally, other possible medical applications of HMDs in diagnostic and screening will be explored since the vestibular system is also involved in other common conditions such as motion sickness and migraine.
SURF-IOT Fellows not attending the final symposium:
PAINBUDDY: USING MOBILE APPLICATIONS TO ASSIST AND TRACK HOME-BASED THERAPY FOR CHILDREN
Fellow: Tyler Stevens
Mentors: Dr. Sergio Gago, Dr. Michelle Fortier
Many of the 12,000 children diagnosed with cancer in the United States each year experience some form of chronic pain and other severe symptoms during and after treatment. Long-term, at-home treatment can be challenging – physicians are reluctant to prescribe potentially addictive pain relievers to their patients, and parents often underestimate the pain their child experiences. This project attempts to provide an extremely low-cost alternative to medication with an emphasis on personalization of care. We are building an Android application featuring a twice-daily survey of the patient's symptoms. After each survey has been completed, the patient is guided as needed through in-application relaxation and pain management exercises to ease their specific symptoms. To motivate the patient to use the application, the application features a customizable set of avatars that unlock as the patient completes more surveys. We have also created a website for healthcare providers to monitor their patients' symptoms over time in order to guide treatment. Finally, we have designed the application to measure statistics about how the patient uses the application, including the amount of time that the patient spends in each section and what avatars and accessories the patient prefers. We have released a beta version of the application to the Children's Hospital of Orange Country, where a small trial is being conducted. We will use the results of the trial and patient feedback to prepare the application for a larger trial planned for 2016.
DESIGN AND DEVELOPMENT OF AN MHEALTH APPLICATION: DISCOVERING THE SPECIFIC NEEDS AND
VALUES OF LATINA BREAST CANCER SURVIVORS
Fellow: Nicole Ulgado, Public Health Sciences
Faculty Mentor: Dara Sorkin, Ph.D., UC Irvine Department of Medicine
CoMentors: Yunan Chen, Ph.D. & Alfred Kobsa, Ph.D. UC Irvine Department of Informatics
Breast cancer is the most common cancer diagnosed in Latinas in the US and pproximately only 32% of breast cancer survivors engage in the recommended level of physical activity. Women with breast cancer face difficulties in coping with changes in lifestyle and health. With the current growth of healthcare costs in mind, utilizing consumer-facing health technologies as personal health management tools can provide a cost-effective solution. The technical capabilities of consumer-facing health technologies such as tablets, personal digital assistants, and mobile phones, have afforded users ubiquitous storage of health information of which may provide beneficial suggestions and more importantly, the awareness of previously unknown habits. This study aimed to generate a technological solution for improving the lifestyle behaviors of Latina breast cancer survivors. Data was collected from 48 Latina breast cancer survivors (age: 33-85 years old, mean: 56.2 [SD: 11.3], years since diagnosis: range: 113 years, mean: 5.4 [SD: 3.6]) through surveys and focus groups and analyzed to determine behavioral patterns along with needs, personnel, and values. It was found that 81% of participants owned a smartphone and of these, about 57% used an iPhone. Over 63% considered themselves having some sophistication when using mobile technology, but over 43% never used the internet to look up health-related information on a device. Conclusions from analyses were utilized to discover ways to integrate findings into a motivational mobile application. Future analyses of user-testing will examine the application’s efficacy and determine ways in which its design can be improved.