The Wireless Health Institute at UCLA is currently recruiting engineering undergraduates for our 2017 Summer Undergraduate Scholars Program. Applicants with coding experience are highly encouraged to apply.
Advances in engineering and computer science are enabling the design of powerful home and mobile technologies that can augment functional independence and daily activities of people with physical impairments, disabilities, chronic diseases, and the accumulative impairments associated with aging. The Wireless Health Institute at UCLA is a collaboration between UCLA Schools of Medicine, Nursing, Engineering and Applied Sciences, Business, the Clinical Translational Science Institute for medical research, and the Ronald Regan UCLA Medical Center aimed at developing new technology essential to the next generation of health care. We encourage undergraduates to get involved with our Institute by participating in our Summer Undergraduate Scholars Program. This 8 – 10 week program provides undergraduates the opportunity to gain handson research experience at WHI labs; to work with WHI faculty, staff and graduate students; and to participate in research, professional development, and social activities with other WHI undergraduate scholars.
EXAMPLES OF RESEARCH PROJECTS:
Acoustic Environment Classification
Classification of different types of sounds is important for establishing the physical context (e.g., indoors, outdoors, busy, quiet). This can be useful in studies of stress or to determine context for other medical data (e.g., physical activities). The students will first learn how to analyze acoustic data (speech, music, and bird songs for example) and to classify them, showing how different recording devices (highquality microphones versus those recorded by a telephone for example) can affect the quality of the recorded sounds. Students will also learn how to analyze sounds and will realize that different sounds have different frequency spectra (distribution of energy versus frequency) and/or timedomain properties (jitter, shimmer, modulation, etc.) Once they have a good understanding of the signals’ properties, they will collect data with careful groundtruth through a set of daily activities, and determine which classification techniques to employ.
Internet of Things for Wireless Health
Advances in commercial technologies have enabled embedded systems to have full Linux capability together with seamless connections to cloud computing via radios. The Intel Edison platform is low cost and is compatible with a wide set of sensors such as those that interface with the popular Arduino platform. The project will involve combining such platforms with wearable sensors to provide immersive home environments that enable health monitoring with a diverse set of sensing modalities and enabling complete end to end systems with open architectures.
Reliable Inference of Upper Limb Activities
Inference of activities of the upper limbs using wearable inertial sensors is much more difficult than the lower limbs. There are fewer opportunities to compensate for drift using windows of zero velocity, and the motions themselves have many more variations. Yet the quality and frequency with which motions necessary for daily living take place are the key indicators of the success of rehabilitative therapy. We propose beginning with development of algorithms for the easier class of repetitive practice exercises, and proceeding to more free form motions by making use of user specific constraints and the presence of additional sensors at selected locations in the environment.
Data Collection and Predictive Analytics
The WHI and ER lab in particular benefit from integrating and harnessing various engineering techniques to tackle medical problems. The domain of our research can be highlighted from signal processing to machine learning, from developing software and applications to programming FPGAs. In our lab we can offer interns supervision and guidance to work on developing health related applications that can be installed on smart wearables, such as smart watches. We can also involve interns in designing and developing health monitoring systems including implementation, data collection, data processing and troubleshooting of such systems.
Energy aware Inferencing of Rich Contexts and Behaviors
Crucial to mHealth are software applications and services that run on mobile smart phones and watches and make realtime inferences of user’s contexts and behaviors from sensory information. At the core of such applications are embedded statistical machine learning algorithms for inferring states and detecting events in realtime, that involve feature computation, online model personalization, and classification. The limited energy and computational resources on such devices make continual background inference of rich contexts very challenging. We will explore the use of external context information supplied by objects in the environment (e.g., for location or identification), lower energy algorithms that successively approximate state as needed and maximal use of onboard sensors to avoid expensive radio communications.
US Citizen or US Permanent Resident.
3.2 GPA minimum.
Continuing undergraduate enrolled in fall.
Underrepresented minority, women, and those with coding experience are highly encouraged to apply.
June 25, 2017 – August 18, 2017 (8 week program).
June 25, 2017 – September 1, 2017 (10 week program).
Length of program based on award and your availability.
Move-in & Mandatory Program Orientation will be held on Sunday, June 25, 2017.
APPLICATION INFORMATION & LINKS: