Summer 2014

Stethoscope Does Double Duty


Electrical and Computer Engineering faculty Mounya Elhilali, Jim West, and their students have built a high-tech stethoscope to address two problems: ambient noise and untrained users.

Equipped with two microphones, the stethoscope records internal and external sounds and can then effectively cancel out the “noise” from the relevant data. West and Elhilali have also developed a data analysis program, soon to be added to the stethoscope, to identify the kinds of irregularities in lung sounds that require further examination by a doctor.

Ultimately, the researchers believe the stethoscope could improve health care in locations where people have minimal medical training. The stethoscope is currently being field-tested in Africa by Eric McCollum and researchers from Hopkins’ Bloomberg School of Public Health, and industry has expressed interest in commercializing the technology.

Talk to Your Toaster


Thanks to cloud-based speech recognition, we can use Siri or Google Voice to tell our phones to place calls, send texts, or even search for the nearest pizza joint. But what if you want to use voice commands with non-networked devices, like your washing machine or microwave oven? Aren Jansen, a research professor in the Center for Language and Speech Processing (along with graduate student Keith Kintzley and Hynek Hermansky, the Julian S. Smith Professor) wants to make that possible.

The solution Jansen and his colleagues now are patenting uses simple arithmetic to search a limited vocabulary of spoken words to provide non-cloudbased speech translation. And their technology requires so little memory that it easily could run on the kinds of microprocessors that are part of just about every electronic device made today.

So Long, Cardiac Cath?


Due to the limits of current diagnostic testing, more than 1 million unnecessary cardiac catheterizations are performed each year in the United States, according to In Vivo: The Business and Medicine Report, July/August 2011. In addition to subjecting patients to unnecessary medical risks, these procedures come at a hefty annual pricetag of $10 billion.

With their startup, HeartMetrics, Mechanical Engineering’s Rajat Mittal and Albert Lardo, an associate professor of biomedical engineering at Johns Hopkins’ School of Medicine, are poised to radically disrupt current practices in coronary angiography by using high-definition CT scans as an effective and noninvasive means to determine the severity of blockage in an artery.

New, highly accurate CT scanners have recently made visible the differences in both the dispersion patterns of CT contrast as well as concentration gradients between normal and partially blocked arteries. Initial studies correlating these gradients to cardiovascular disease, however, were disappointing.

Using fundamental principles of fluid mechanics, Mittal and Lardo figured out that these contrast gradients are just one piece of the puzzle. Together, they developed an algorithm to decode the relationship between contrast gradients and the severity of the arterial blockage. The researchers are now in the process of validating and testing their technology for commercialization.

In Impact