Associate Professor of Biomedical Engineering
Research Interests: Control, Estimation, and System Identification of Biomedical Systems, Data Science and Computational Methods for Biomedicine, Biomedical and Neural Signal Processing, Wearable Computing, Physiological Modeling & Cyber Physical Systems
Massachusetts Institute of Technology
Postdoctoral, Brain and Cognitive Sciences
Massachusetts Institute of Technology
Ph.D., Electrical Engineering and Computer Science
Massachusetts Institute of Technology
S.M., Electrical Engineering and Computer Science
University of Maryland
B.S. (Summa cum Laude), Electrical Engineering (Honors Program)
- MIT Technology Review Innovator Under 35, 2020
- National Science Foundation CAREER Award, 2020
- Junior Faculty Research Excellence Award, Cullen College of Engineering, University of Houston, 2020
- Teaching Excellence Award, Cullen College of Engineering, University of Houston, 2020
- Featured in IEEE Women in Engineering Magazine as a 'Woman To Watch', 2020
- Selected for the National Academy of Engineering's Frontiers of Engineering Symposium, 2019
- IEEE-USA’s New Face of Engineering, 2016
- National Science Foundation Graduate Research Fellowship, on tenure 2009-2012
- Massachusetts Institute of Technology Graduate Fellowship in Control, 2008
- Department of Electrical and Computer Engineering Chair’s Award, University of Maryland, 2008
- Phi Kappa Phi Honor Society, Inducted in 2008
- Tau Beta Pi, The Engineering Honor Society, Inducted in 2008
- Eta Kappa Nu, The Honor Society of IEEE, Inducted in 2008
- University of Maryland President's Scholarship, 2006-2008
Closed-loop tracking and regulation of emotional valence state from facial electromyogram measurements
Emotions directly influence the way we think and interact with others in different situations. Additionally, when it interferes with rationality in our decision-making or perception, emotions play an undeniable role.
From another perspective, studies from different fields indicate that human emotions could be impacted by any changes in the environment. Thus, having a solid understanding of the dynamics of emotions could benefit individuals in multiple ways: from the enhancement of therapeutic solutions and maintaining well-being to the prevention of unpleasant emotions and inducing favorable ones. Decoding emotions has been an ongoing research among researchers, dictating joint efforts from behavioral, physiological, and computational angles.
According to Russell's circumplex model, emotions are defined in a two-dimensional arousal-valence circular space. Compared to the arousal index, which designates the severity of the emotional state, internal valence levels determine the quality of emotions. In a high arousal mood, feeling excited or happy is opposed to nervousness and feeling stressed. On the flip side and in low arousal conditions, staying calm stands on the opposite side of feeling bored. Hence, feeling pleasant, which refers to the positive valence, is one of the key factors in one’s health condition and could result in a better quality of life.
In a new piece of research, a team led by Rose Faghih, Associate Professor of Biomedical Engineering, follows the goal of establishing a mechanism to track an individual's emotional valence level and regulate it in a closed-loop framework. Since the emotional valence state is hidden and cannot be measured directly, the authors need to measure it indirectly.
The use of wearable devices and designing decoders to gain insight into the internal brain states provides a good alternative to directly measuring brain activity. Inspired by recent advances in wearable technologies which could measure various physiological signals from humans in the loop, there exists possibilities to track emotional valence state indirectly and in a more practical manner. Among different physiological signals, it has been shown that facial electromyography (EMG) carries valuable information about the internal valence state. According to various research, facial EMG measurements of the zygomaticus muscle (zEMG) give the most distinct indicator of emotional valence state. In fact, changes in the zygomaticus muscle, which could be captured via wearables, could provide researchers with valuable information regarding one’s valence level.
In this research, the authors employ a mathematical framework to relate changes in zEMG signal for inferring the internal valence state. They extract corresponding features from the zEMG signal and estimate the internal valence state. Taking advantage of tracking of emotional valence state in a systematic way, they then propose a knowledge-based control architecture to close the loop and keep the estimated valence state within a favorable range. Within a closed-loop approach, biomarkers are collected to provide information regarding the individual's valence state to then automatically adjust the actuators in response to the current emotional state.
In silico results verify that, by making use of physiological data, the proposed controller could effectively regulate the estimated valence state. Ultimately, it is envisioned that the future outcomes of this research will include supporting alternative forms of self-therapy by using wearable machine interface architectures capable of mitigating periods of pervasive emotions and maintaining daily well-being and welfare. This study is the very first closed-loop control framework for regulating emotional valence state using facial muscle activity. This research suggests an important new step toward new clinical applications and the self-management of mental health. In the practical implementation of the proposed architecture, a wearable device would collect zEMG signal, and the appropriate practices (e.g., listening to music, breathing exercises) would be suggested to maintain the emotional valence levels within the favorable range.
Faghih, who joined NYU earlier in 2022, has been integrating research into her classroom teaching. The research covered in this Frontiers in Computational Neuroscience paper started as a course project in Faghih’s “State- Space Estimation with Physiological Applications” course at the University of Houston. In this class students had the opportunity to use advanced methods learned in the course to analyze physiological data in hands-on biomedical engineering projects designed by Faghih. Student authors Luciano Branco and Arian Ehteshami were graduate students in the course, and Hamid Fekri Azgomi, who recently completed his PhD under the supervision of Faghih, was a PhD student mentor for the project.
- Rose Faghih
In the modern world, any challenge might be a source of stress. The fast-paced life has the potential to induce emotional and cognitive stress. Feeling overwhelmed, anxious, and agitated are among the symptoms associated with high levels of cognitive stress. Conversely, loss of cognitive engagement might also prevent individuals from following their goals. A low level of positive stress, which is also called eustress, might cause memory problems, lack of motivation, and poor concentration. It can also negatively affect individuals’ productivity in the workplace. According to multiple reports from “Mental Health America”, the percentage of people with depression and anxiety is growing in the United States. Thus, it is crucial to establish a mechanism for tracking one's cognitive stress level and keeping it within a favorable range.
Influenced by the recent advances in wearable technologies and inspired by the fact that skin conductance signal carries important information regarding one’s internal arousal state, there could be practical approaches to handle cognitive stress levels in daily life. In response to any environmental stimuli, the human brain controls the autonomic nervous system in multiple ways. As changes in electrodermal activity (EDA) signals are impacted by the changes in sweat glands activation, variations in skin conductance signal can be tracked in real-time using wearable devices.
In order to eventually enable wearable technologies to track and regulate one's internal stress state, the lab of Rose Faghih has pioneered methods for estimating the internal arousal state by monitoring one's EDA, i.e., fluctuations in their skin conductivity. They have developed a mathematical framework to model the internal arousal state and relate it to the changes in skin conductance signal (i.e., a signal that can be measured via wrist-worn wearable devices). Moreover, to estimate the hidden arousal state, they have developed methods that analyze the EDA signal and extract the underlying arousal events that activate sweat glands.
Systematic modeling of the internal arousal state enables developing controllers to close the loop and regulate the estimated stress state. In this regard, well-established control techniques (i.e., linear quadratic regulator and model predictive controller) can be applied. In the process of finding the optimal control policy to regulate the stress state within these approaches, a corresponding cost function needs to be minimized. Hence, the performance of these control systems highly depends on the selection of corresponding objective functions.
As the researchers are dealing with a human-in-the-loop problem, the process of finding the appropriate objective functions becomes more complicated. To overcome this challenge and enhance the performance of the closed-loop system, Faghih and Fekri Azgomi, who recently completed his PhD under the supervision of Faghih, developed a supervised controller that includes a knowledge-based supervised layer that adjusts the objective function and tunes parameters in real-time. In this supervised architecture, a fuzzy system is utilized to continuously update parameters in the objective function to adjust environmental variations. The results verify the effectiveness of proposed architecture in keeping the estimated stress state within a desired range by designing a supervised layer on top of the optimal control formulation.
Furthermore, the proposed supervised control architectures provide an excellent setting to incorporate the relevant medical expertise to enhance the closed-loop system. These novel supervised control approaches could be further expanded to deliver adaptive and robust closed-loop characteristics to address inter- and intra-subject variations.
There also exists great potential in implementing the supervised control architectures in other closed-loop applications such as adaptive deep brain stimulation for the treatment of neurodegenerative brain disorders and automating medication intake in a wide range of diseases. Once one models and tracks one's internal states, utilizing supervised control methodologies would result in enhancing the closed-loop treatment practices in a more personalized manner.
- Rose Faghih
This research included Rose T. Faghih, Associate Professor of Biomedical Engineering at NYU Tandon, and researchers from the University of Houston, the Louis Bolk Instituut, and the Weill Institute for Neurosciences.
According to the CDC, obesity prevalence in the United States was 42.4 percent in 2017–2018. Obesity prevalence in the United States climbed from 30.5 percent to 42.4 percent between 1999 and 2018. Obesity has been linked to cardiovascular disease, stroke, Type 2 diabetes, and a variety of malignancies. These are some of the most common causes of death that can be avoided. In 2008, the yearly medical cost of obesity in the United States was estimated to reach $147 billion.
It is important to understand the root cause of obesity in terms of hormonal imbalance. To regulate food intake, the brain responds to signals from fat (adipose) tissue, the pancreas, and the digestive tract. These instructions are transmitted via hormones like leptin, insulin, and ghrelin, as well as other small molecules. Among them, leptin is a signaling hormone that is essential to signal the brain in the suppression of appetite. It regulates food intake, metabolism, energy expenditure, and body weight. Thus, leptin regulation is known to be closely related to diseased conditions such as obesity.
In a new study from researchers at NYU Tandon and other schools, the group analyzed the behavior of two neuroendocrine hormones, leptin and cortisol (a stress hormone), in a cohort of patients with obesity. They used a system theoretic approach that can accurately estimate the internal secretion patterns, timings, amplitudes, number of underlying hormone secretory pulses, infusion, and clearance rates of hormones in patients with obesity by only measuring the 24-hour blood assay of their hormones.
The findings suggest a method for mathematically modeling both leptin and cortisol hormones to characterize how they interact as part of a larger system. Because the relationship between leptin and cortisol hormones is complex, the new results and projections will help us understand how these hormones work together to keep the body in a state of homeostasis.
The researchers demonstrated a negative relation between leptin and cortisol secretion, based on a statistical test called the Granger causality test among the patients with obesity. These results indicate that increases in cortisol are prospectively associated with reductions in leptin, suggesting a negative inhibitory relationship in 14 out 18 obese women investigated. Reduced leptin may result in a decrease in satiety and thereby lead to obesity
The model can be a crucial contribution toward the potential development of the next generation of agile closed-loop medical systems related to obesity. Such next-generation closed-loop medical systems will identify deviations from homeostasis and suggest necessary interventions benefiting from regular tracking of hormone secretory events and underlying endocrinological system parameters. This way complex conditions such as obesity can be prevented at the root level resulting in an overall increase in the quality of life and reduction in the total national medical expenditure.
- Rose Faghih