Wearable Acoustic & Vibration Sensing
and Machine Learning for Human Health & Performance
Dr. Omer Inan
November 30, 2022 | 9:00 am CST
Recent advances in digital health technologies are enabling biomedical researchers to reframe health optimization and disease treatment in a patient-specific, personalized manner. This talk will focus on my group’s research in three areas of relevance to digital health: (1) cardiogenic vibration sensing and analytics; (2) musculoskeletal sensing with joint acoustic emissions and bioimpedance; and (3) non-invasive neuromodulation for stress. Our group has extensively studied the timings and characteristics of cardiogenic vibration signals such as the ballistocardiogram and seismocardiogram, and applied these signals for cuffless blood pressure measurement, heart failure monitoring, and human performance. We have also leveraged miniature contact microphones to measure the sounds emitted by joints, such as the knees, in the context of movement, and have examined how these acoustic characteristics are altered by musculoskeletal injuries and disorders (e.g., arthritis). Finally, we have developed non-pharmacological treatment paradigms for posttraumatic stress disorder (PTSD) based on non-invasive vagal nerve stimulation, and have performed extensively validation of this approach with collaborators in psychiatry and radiology. We envision that these technologies can all contribute to improving patient care with lower cost.

Dr. Omer Inan is Professor and Linda J. and Mark C. Smith Chair in Bioscience and Bioengineering in the School of Electrical and Computer Engineering, and Adjunct Associate Professor in the Coulter Department of Biomedical Engineering, at Georgia Tech. He received his BS, MS, and PhD in Electrical Engineering from Stanford in 2004, 2005, and 2009, respectively. From 2009-2013, he was the Chief Engineer at Countryman Associates, Inc., a professional audio manufacturer of miniature microphones and high-end audio products for Broadway theaters, theme parks, and broadcast networks. His research focuses on non-invasive physiological sensing and modulation for human health and performance, and is funded by DARPA, NSF, ONR, NIH, CDC, and industry. He has published more than 300 technical articles in peer-reviewed international journals and conferences, and has twelve issued patents. He has received several major awards for his research including the NSF CAREER award, the ONR Young Investigator award, and the IEEE Sensors Council Early Career award. He also received an Academy Award for Technical Achievement from The Academy of Motion Picture Arts and Sciences (The Oscars). He is an Elected Fellow of the American Institute for Medical and Biological Engineering (AIMBE). While at Stanford as an undergraduate, he was the school record holder and a three-time NCAA All-American in the discus throw.
Intelligent Critical Care:
Opportunities & Challenges
Dr. Parisa Rashidi
October 28, 2022 | 9:00 am CST
In recent years, we have witnessed a rapid surge in building intelligent health systems. Artificial intelligence and machine learning techniques are central to all these systems and constitute a major step towards developing more intelligent healthcare solutions. These techniques not only make it possible to process and transform data into actionable knowledge, but also facilitate decision making and reasoning. This talk will discuss the rise of intelligent health systems in patient monitoring and will explore the challenges and opportunities in this area.

Dr. Parisa Rashidi is the foudning co-diretcor of the Intelligent Critical Care Center (IC3) at the University of Florida (UF) and an associate professor at the J. Crayton Pruitt Family Department of Biomedical Engineering (BME). She is also affiliated with the Electrical & Computer Engineering (ECE) and Computer & Information Science & Engineering (CISE) departments. Her research aims to bridge the gap between machine learning and patient care.
Dr. Rashidi is a National Science Foundation (NSF) CAREER awardee, the National Institute of Health (NIH) Trail Blazer Awardee, Herbert Wertheim College of Engineering Leadership Excellence Awardee, Herbert Wertheim College of Engineering Assistant Professor Excellence Awardee, and a recipient of the UF term professorship. She is also a recipient of UF’s Provost excellence award for assistant professors; with more than 500 tenure-track assistant professors at UF, Dr. Rashidi is one of only 10 to receive this award. She was invited by the National Academy of Engineering (NAE) as one of only 38 outstanding US engineers under 45 to participate in the EU-US Frontiers of Engineering (FOE) Meeting. To date, she has authored 170+ peer-reviewed publications. She has chaired six workshops and symposiums on intelligent health systems and has served on the program committee of 20+ conferences. Dr. Rashidi’s research has been supported by local, state, and federal grants, including awards from the National Institutes of Health (NIBIB, NCI, and NIGMS) and the National Science Foundation (NSF).
Population Coding
in the Cerebellum
A slow sensory system presents major problems for movement control. Yet, despite this shortcoming the healthy brain composes exquisite movements. Textbooks posit that this remarkable ability is due to the cerebellum, a structure that learns to predict sensory consequences, thus overcoming time delays. However, cerebellar neurons fire in patterns that do not correspond well with movements. Thus, the language with which the cerebellum expresses its predictions has remained a mystery. The idea that we have explored is that in the cerebellum, the fundamental unit of computation may not be a single neuron, but a group of neurons that share the same teacher. In this analogy, the teacher is the inferior olive, organizing the students (Purkinje cells) into groups. To test this idea, we have measured activity of neurons in macaques and marmosets and found that while activity of individual neurons is difficult to decipher, activity of a group of neurons that shares the same teacher is a rather precise predictor of the ongoing movement, particularly during deceleration and stopping.
Dr. Reza Shadmehr
September 30, 2022 | 9:00 am CST

Dr. Reza Shadmehr is a professor of biomedical engineering and neuroscience at the Johns Hopkins University School of Medicine. His research focuses on understanding how the human brain perceives the world, how it learns and how it controls our movements. Dr. Shadmehr also serves as co-director of the Biomedical Engineering Ph.D. program at Johns Hopkins University. Dr. Shadmehr received his undergraduate degree in electrical engineering from Gonzaga University. He earned a master’s degree in biomedical engineering and a Ph.D. in computer science (robotics) from the University of Southern California. Dr. Shadmehr completed the McDonnell-Pew post-doctoral fellowship at MIT and joined the Johns Hopkins faculty in 1995.
He has many published works including two books, The Computational Neurobiology of Reaching and Pointing and Biological Learning and Control. Dr. Shadmehr also has two patents filed.
Patient-specific Modeling
for Virtual Treatment Planning in Cardiovascular Disease
Dr. Alison Marsden
August 26, 2022 | 9:00 am CST

Dr. Alison Marsden is the Douglass M. and Nola Leishman Professor of Cardiovascular Disease in the Departments of Pediatrics, Bioengineering, and, by courtesy, Mechanical Engineering at Stanford University. She is a member of the Institute for Mathematical and Computational Engineering. From 2007-2015, she was a faculty member in Mechanical and Aerospace Engineering at UCSD. She graduated with a BSE degree in Mechanical Engineering from Princeton University in 1998, and a PhD in Mechanical Engineering from Stanford in 2005. She was a postdoctoral fellow at Stanford University in Bioengineering from 2005-07. She was the recipient of a Burroughs Wellcome Fund Career Award at the Scientific Interface in 2007, an NSF CAREER award in 2011. She was elected fellow of AIMBE and SIAM in 2018, the APS DFD in 2020, and BMES in 2021. Her research focuses on the development of numerical methods for cardiovascular blood flow simulation and application of engineering tools to impact patient care in cardiovascular surgery and congenital heart disease.
Cardiovascular disease is the leading cause of death worldwide, with nearly 1 in 4 deaths caused by heart disease alone. In children, congenital heart disease affects 1 in 100 infants, and is the leading cause of infant mortality in the US. Patient-specific modeling based on medical image data increasingly enables personalized medicine and individualized treatment planning in cardiovascular disease patients, providing key links between the mechanical environment and subsequent disease progression. We will discuss recent methodological advances in cardiovascular simulations, including (1) uncertainty quantification to assess reliability of simulation predictions, and (2) a unified finite element formulation for fluid structure interaction and fluid solid growth simulations. Clinical application of these methods will be demonstrated in two clinical applications: 1) virtual treatment planning in pediatric patients with peripheral pulmonary stenosis, and 2) prevention of vein graft failure after coronary bypass graft surgery. We will briefly discuss our open source SimVascular project, which is available to the scientific community (www.simvascular.org). Finally, we will provide an outlook on recent successes and challenges of translating personalized simulation tools to the clinic.
Seismocardiography
Genesis, and Utilization of Machine Learning for Variability Reduction and Improved Cardiac Health Monitoring
Seismocardiography (SCG) is the measured chest surface vibrations resulting from the cardiac activity. Although SCG can contain information that correlates with cardiac health, its utility is limited by a lack of understanding of the signal genesis and variability that can mask subtle SCG changes. Research presented in this talk address the genesis of SCG via Finite Element Modeling (FEM) of cardiac related chest vibrations and reduction of SCG variability using unsupervised machine learning (ML). The effects of cardio-pulmonary interactions on the SCG variability will be also analyzed. FEM analysis suggested that ventricular movement is a primary source of SCG. We will show that unsupervised ML helps reduce the SCG variability by clustering SCG beats into groups with minimal intra-cluster heterogeneity and unveiled consistent relations with the respiratory phases and SCG morphology. In our research, the longitudinal SCG measurements from reduced ejection fraction heart failure (rEFHF) patients were also utilized to predict patient readmission. Here, many time- and frequency-domain SCG features were extracted from clustered SCG beats. Using supervised ML, high classification accuracies were achieved suggesting high SCG utility for monitoring rEFHF patients and possibly other heart conditions.
Dr. Peshala Gamage
March 21, 2022 | 3:30 pm CST
.png)
Dr. Peshala Gamage is an Assistant Professor in the Department of Biomedical and Chemical Engineering and Sciences at Florida Institute of Technology. Prior to his current role, he worked as a postdoctoral researcher in Florida Tech. He received his Ph.D. degree in Mechanical Engineering from the University of Central Florida in 2020 and received his M.Sc. degree in Aerospace Engineering from the same university in 2017. His current research interests lie in the areas of physiological signal processing, machine learning, biomedical acoustics, and computational modeling.
Daily Variations of
Brain Connectivity
Patterns:
A Graph-based Analysis
Dr. Farzad V. Farahani
April 25, 2022 | 3:30 pm CST

Dr. Farahani is a Postdoctoral Researcher in the Department of Biostatistics at Johns Hopkins University since Fall 2020. He earned his Ph.D. in the Computational Neuroergonomics track of the Industrial & Systems Engineering program at the University of Central Florida in 2020.
His primary research interest is to better understand the brain as a complex system in relation to behavioral performance, as well as how information is integrated across functionally specialized neural units that reside in spatially disparate brain regions. His current work focuses on analyzing connectivity patterns in the human brain using computational models such as graph theory and machine learning, as measured with functional and anatomical neuroimaging methods.
Most living organisms express a rhythmic cycle across a 24-hour period (circadian rhythm) that controls several physiological processes such as sleep–wake patterns, metabolic activity, and body temperature, as well as various brain functions such as attention, decision making, motor activity, and visual detection tasks. Furthermore, individuals have biologically different inclinations for when to sleep and when they are at their highest alertness and energy level, which are referred to as chronotypes. In this study, using graph-based knowledge and noninvasive imaging modalities such as functional MRI (fMRI), we examined the effect of both time of day and the individual’s chronotype on whole-brain network organization. The findings provide insight into daily variations in resting-brain networks, reflecting the universal effect of time-of-day on neural functional architecture when designing experiments. The findings also indicate the need to control for circadian typology, which could influence experimental results in neuroimaging studies.