Laypeople may think of engineering as being more about numbers and materials than human beings. But thats not the case, especially with bioengineering. Its a field that synthesizes engineering techniques with the biological sciences. Electrical and computer engineering professor Pushkin Kachroo and his department have been committed to expanding bioengineerings reach since his arrival at UNLV more than 10 years ago. With the creation of the universitys School of Medicine, the opportunities for collaboration are growing.
Its a common-sense link that health, medicine and engineering, should be connected together, Kachroo said. Thats where health is going.
Thanks to the new Howe Fellowship in Bioengineering at UNLV, the work of two engineering graduate students is getting a boost in showcasing the ways engineering can materially benefit human health and well-being.
Lina Chato was always good at math growing up in her native Iraq, but she found her real passion when her school opened a new computer lab. Fellow students showed her some simple programs, let her borrow books, and before long she had written her first program about computer-aided learning for electrical circuit design. She was only 13 years old.
Chato received a bachelors degree in computer engineering and a masters degree in mechatronics engineering from the University of Technology, Baghdad. She joined the faculty and published three papers, but when the country became unstable due to the war, she fled with her family to the United States. She settled in Las Vegas, where two of her sisters already lived, and applied to UNLV's College of Engineering. Chato worked as a research assistant before beginning her doctorate program.
I came as a refugee, without any source of money, she said. This first opportunity was really appreciated and important to start my Ph.D. study.
Chato became interested in using machine learning to try to better analyze MRI images of brain tumors. Her work revolves around developing models to predict survival time for Glioma tumor patients. These types of tumors represent nearly three quarters of all malignant tumors.
We still dont know how these tumors behave, she said. The behavior of tumors is an important factor in predicting survival. This model can describe how a brain tumor develops. If we can know that, we can use this in the treatment stage.
When he graduated from the Las Vegas Academy of the Arts, Jadin Tredup wasnt sure what he wanted to do pursue music or study math. So he entered UNLVs then newly created entertainment engineering program.
It seemed like a pretty good blend of the two, Tredup said.
At UNLV his focus shifted away from music toward mechanical engineering, and then again toward electrical engineering. His journey took him from robotics to machine learning and artificial intelligence, to the problem of applying brain wave signals.
Now, having just finished his masters degree, hes working on a model to help patients he characterizes as having profound intellectual and multiple disabilities people with very severe and limited cognitive and motor functioning and an inability to communicate verbally.
The idea is, because our bodies carry so much more information beyond what we can produce in words, if we can sense all these physiological signals we can then translate them using AI and machine learning into a language and vocabulary for people, he said.
Tredups work leverages a theory about how the environment impacts how we communicate. For instance, EEGs can measure brain waves and translate them into emotional states; sensors can measure galvanic skin response (sweat, basically, an indicator of emotional state); and eye tracking can measure what patients are paying attention to. They can detect nearly instantaneous changes, allowing researchers to deduce, for example, that a barking dog might be causing anxiety. You can translate the data into a basic language, like Im feeling anxious because of the dog barking.
The next step is to develop a needs assessment for a few specific patients at a care facility, then further develop the algorithm.
No one model is easily applicable person-to-person, Tredup said. We have to create a generalizable model.
Tredup began his doctorate program at UNLV this fall. (He is the fourth in his family after his parents and brother to graduate from UNLV.) The Howe Foundation Fellowship will afford him more time to concentrate on research instead of juggling full-time research and a full-time job. Similarly, Chatos award will help her focus more energy on her research and spend less time as a teaching assistant.
Working with algorithms is cool, but it doesnt mean much unless it has applicability to the world around it, said Tredup. I can use everything Ive learned and help better peoples lives.
Read the rest here:
Building the Body | News Center - UNLV NewsCenter
Recommendation and review posted by G. Smith
