The turning point, as Sunwoo has described it, came from seeing neuromodulation in action. Deep brain stimulation, in which implanted electrodes deliver carefully timed electrical pulses to targeted nuclei such as the subthalamic region, can dramatically reduce the tremors, rigidity, and motor fluctuations that define advanced Parkinson’s disease. For a young student watching patients regain control of their own bodies, the implication was electrifying: if electricity delivered to a precise brain target could restore function, then perhaps the circuitry of the brain itself could be mapped, understood, and ultimately reconstructed. It was a question that would define the first major decision of his scientific life.

That decision was to switch majors and pursue a doctorate in biomedical engineering, training at Cornell University and then the University of Southern California. His doctoral work concentrated on noninvasive physiologic monitoring, brain oxygenation, and the autonomic nervous system, with a strong emphasis on studies involving actual patients rather than isolated laboratory models. This clinical orientation would prove decisive. Working with human subjects teaches a particular kind of humility, Sunwoo has noted, because the brain in all its complexity resists the tidy reverse-engineering fantasy that first drew him to neuroscience. Understanding even a small slice of neural function, he realized, demands starting from the foundational elements — the measurable physiological signals that reveal how the body, and the brain within it, is actually behaving.

So instead of abandoning the dream, Sunwoo scaled it down and grounded it. Noninvasive physiologic monitoring became his new mission, and the brain remained very much at its center. After completing his PhD, he joined the Athinoula A. Martinos Center for Biomedical Imaging at Massachusetts General Hospital, one of the world’s leading facilities for advanced imaging and optical monitoring of human physiology. There he trained in diffuse correlation spectroscopy, a technique that has quietly become one of the most promising tools for measuring cerebral blood flow without needles, radiation, or surgery.

Diffuse correlation spectroscopy, or DCS, belongs to the broader family of near-infrared spectroscopy methods, but it does something distinct. Near-infrared light, at wavelengths between roughly 650 and 900 nanometers, penetrates several centimeters through scalp, skull, and other tissue. Conventional near-infrared spectroscopy measures changes in the absorption of that light to estimate the balance of oxygenated and deoxygenated hemoglobin — in other words, oxygenation. DCS goes a step further by exploiting a quantum optical phenomenon: when coherent laser light scatters off moving red blood cells, the reflected light fluctuates in a pattern that encodes the speed of those cells. By analyzing these intensity fluctuations with a correlation algorithm, researchers can derive an index of blood flow in the microvasculature beneath the sensor. The result is a continuous, bedside measurement of perfusion in tissue that would otherwise require CT perfusion scans, MRI, or invasive probes to assess.

For Sunwoo, the appeal of this technology found its natural application in one of medicine’s most vulnerable populations: preterm infants in the neonatal intensive care unit. These patients are non-verbal, their physiology is extraordinarily labile, and their brains are in a period of explosive development that makes them exquisitely sensitive to disturbances in blood flow and oxygen delivery. Continuous monitoring is not a luxury in this setting; it is a clinical necessity. Yet standard NICU monitoring focuses heavily on heart rate, oxygen saturation, and blood pressure — systemic signals that may not reflect what is actually happening inside the cerebral circulation. A bedside optical sensor that tracks cerebral blood flow directly offers a way to close that gap, and Sunwoo’s work at the Martinos Center placed him at the intersection of optical engineering and neonatal critical care.

His published research illustrates the promise concretely. In a 2022 study in Scientific Reports, Sunwoo and colleagues applied DCS blood flow monitoring to extremely low gestational age newborns, examining how cerebral perfusion patterns relate to the vulnerability of these infants to intraventricular hemorrhage. Intraventricular hemorrhage, bleeding into the brain’s ventricular system, remains one of the most feared complications of extreme prematurity, and its onset has long been linked to unstable fluctuations in cerebral blood flow. By measuring flow continuously at the bedside, the study sought to identify which infants are most at risk and when that risk peaks — information that could eventually guide interventions to stabilize perfusion before injury occurs.

More recently, Sunwoo has turned his attention to another dimension of neonatal brain protection: the autonomic nervous system. In a 2026 study published in Pediatric Research, he and his collaborators assessed parasympathetic tone in infants undergoing therapeutic hypothermia for brain injury following neonatal encephalopathy. Therapeutic hypothermia, the deliberate cooling of the body to around 33.5 degrees Celsius, is the standard of care for infants with hypoxic-ischemic encephalopathy, a condition in which oxygen deprivation around the time of birth damages the brain. Cooling slows the cascade of metabolic injury, but it also profoundly alters autonomic regulation, including the vagal control of heart rate. By quantifying parasympathetic tone during cooling and rewarming, the research aims to understand how the autonomic nervous system responds to this therapy and whether those responses carry prognostic information about which infants will recover and which will sustain lasting damage.

Today, as a faculty member in Pediatrics at Harvard Medical School and an investigator at Mass General Brigham for Children, Sunwoo’s laboratory focuses on using physiologic signals to better understand and protect infants at risk of brain injury. His work is supported by the National Institutes of Health, including grants from the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Neurological Disorders and Stroke, along with funding from the Harvard Medical School Office for Faculty Affairs. The trajectory that began with a dream of reverse-engineering the brain has matured into something arguably more achievable and more immediately humane: reading the physiological signals of patients who cannot report their own symptoms, and translating those signals into clinical protection.

There is a broader lesson in this career arc that resonates well beyond one researcher’s story. Big scientific dreams rarely survive first contact with the brain’s staggering complexity in their original form, but they can be redirected into productive channels. The student who wanted to rebuild the brain became the scientist who measures its blood flow, one photon at a time, in infants weighing less than a kilogram. The humility that Sunwoo describes — learning firsthand what it takes to understand the brain — did not extinguish the original ambition. It refined it, grounding a grand vision in the quiet, continuous work of monitoring, measuring, and protecting the most fragile patients in medicine. In neonatal intensive care, where every bit of continuous monitoring matters, that grounded vision may ultimately prove as transformative as the dream that started it all.

Subject of Research: Noninvasive optical monitoring of cerebral blood flow and autonomic function in newborns at risk of brain injury

Article Title: ECI biocommentary: John Sunwoo

Article References: Sunwoo, J. (2026). ECI biocommentary: John Sunwoo. Pediatric Research. https://doi.org/10.1038/s41390-026-05426-1

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05426-1

Keywords: diffuse correlation spectroscopy, near-infrared spectroscopy, cerebral blood flow, neonatal intensive care, preterm infants, intraventricular hemorrhage, therapeutic hypothermia, neonatal encephalopathy, autonomic nervous system, parasympathetic tone, brain injury, biomedical engineering

The research team, based at the University of Delaware, recruited 64 healthy, typically developing children and adolescents between the ages of 7 and 17. Participants were divided into a younger group of 35 children aged 7 to 12 and an older group of 29 adolescents aged 13 to 17. All were non-obese, free of chronic disease, and not taking any medications known to influence blood vessel function or sleep. After fasting for more than six hours and abstaining from caffeine and vigorous exercise for 24 hours, each participant underwent a five-minute electrocardiogram recording while lying quietly in a dimly lit room, breathing spontaneously. The recordings were sampled at 1,000 Hz and carefully cleaned of ectopic beats and artifacts before analysis.

What makes this study unusual is its dual approach. Most pediatric HRV research relies on linear measures—statistical summaries of beat-to-beat intervals in the time domain, or power spectral analysis in the frequency domain. These metrics, such as SDNN (the standard deviation of normal interbeat intervals) and RMSSD (the root mean square of successive differences), are well validated and widely used, but they treat the heart rate signal as if it were a stationary process. In reality, the heartbeat is shaped by multiple interacting physiological systems and fluctuates continuously in response to internal and external stimuli. To capture this complexity, the researchers also applied nonlinear techniques drawn from chaos theory: approximate entropy and sample entropy to quantify the regularity and predictability of the rhythm, detrended fluctuation analysis to probe its fractal organization, and Poincaré plots to visualize the geometry of beat-to-beat dynamics.

The linear results were, on the surface, relatively straightforward. Adolescents showed a longer mean RR interval—896 milliseconds compared with 813 milliseconds in children—reflecting a slower resting heart rate, consistent with the growth of the heart and increased stroke volume that accompany maturation. Yet this slowing did not translate into higher linear HRV. There were no significant age differences in SDNN, RMSSD, or any of the frequency-domain measures. The most notable linear finding was a sex difference: when data were pooled across age groups, boys showed higher SDNN (70 versus 55 milliseconds) and higher RMSSD (83 versus 61 milliseconds) than girls, indicating greater overall variability and stronger parasympathetic modulation in males.

The nonlinear analysis told a more surprising story. Approximate entropy, a measure of how irregular and unpredictable the heart rate time series is, was significantly lower in adolescents than in children—and the effect was driven almost entirely by boys. Adolescent males had an ApEn of 1.07 compared with 1.18 in male children, a difference with a large effect size. In other words, as boys moved through puberty, their heart rate rhythms became more regular and more predictable. Detrended fluctuation analysis reinforced this picture: the intermediate-term scaling exponent DFA α2 was lower in adolescents (0.24 versus 0.30), moving closer to 0.5, the value associated with uncorrelated randomness, and away from 1, the value associated with healthy fractal self-organization.

These findings complicate a long-standing theoretical framework. The concept of ‘loss of complexity,’ proposed by Lipsitz and Goldberger in the 1990s, holds that aging and disease strip physiological signals of their intricate variability, leaving them more rigid and less adaptable. But the authors of the new study point out that this framework was never intended to describe childhood. During development, the relationship between regularity and complexity may not be so simple. An erratic, irregular heart rate pattern in a young child may not represent the same kind of adaptive complexity that it does in an adult. The researchers suggest that ‘irregularity’ and ‘complexity’ may not be synonymous during development, and that the changing demands of puberty—shifting hormones, growing cardiovascular capacity, and maturing neural regulation—may reshape the heart’s dynamics in ways that do not map neatly onto adult models.

The Poincaré analysis added further nuance. This technique plots each RR interval against the one that follows it, producing a scattergram whose dimensions reveal short-term variability (SD1) and long-term variability (SD2). Boys showed higher SD1 than girls when data were pooled across ages (59 versus 43 milliseconds), mirroring the sex difference in RMSSD, to which SD1 is mathematically related. More intriguingly, adolescent girls showed significantly lower SD2 than female children (45 versus 81 milliseconds), hinting that long-term variability may follow a distinct developmental course in females. The correlations between nonlinear and linear metrics were also revealing: SD1 correlated almost perfectly with RMSSD (rho = 0.98), while the fractal scaling exponents correlated negatively with vagal markers such as RMSSD and high-frequency power, suggesting that linear and nonlinear measures capture overlapping but distinct features of cardiac regulation.

The study has important limitations that the authors acknowledge candidly. The cross-sectional design cannot establish causation, and the modest sample size of 64, with sex-by-age subgroups ranging from 11 to 23 participants, limits statistical power. Breathing rate was not quantitatively controlled, which matters because respiration directly influences parasympathetic HRV indices. The five-minute recording paradigm, while validated and practical for pediatric populations, differs from the 24-hour Holter monitoring used in much of the existing literature, which may explain some discrepancies with earlier studies that reported steady age-related increases in linear HRV throughout childhood. Pubertal status was assessed by questionnaire rather than hormonal measurement, leaving the precise role of sex hormones unexplored.

Nevertheless, the study is the first to characterize spontaneous-breathing, short-term nonlinear HRV during childhood development, and its implications reach beyond basic physiology. Reduced nonlinear HRV is a robust predictor of cardiac recovery and mortality risk in adults with heart failure, and the nonlinear structure of the heart rate signal is increasingly viewed as a promising subclinical biomarker. If the developmental trajectory of these nonlinear properties can be mapped precisely, clinicians may one day be able to identify when a child’s cardiac autonomic regulation deviates from the typical path—whether in obesity, sleep disorders, anxiety, or congenital heart disease—long before overt symptoms appear. The authors call for longitudinal and mechanistic studies to determine what drives the shift toward regularity in adolescence, and how it differs between boys and girls. For now, the message is clear: the developing heart is not simply a smaller version of the adult heart, and its rhythms hold secrets that only nonlinear analysis can reveal.

Subject of Research: Short-term linear and nonlinear heart rate variability in typically developing children and adolescents

Article Title: Characterizing short‐term linear and nonlinear heart rate variability in typically developing children and adolescents

Article References: Matias, A. A., D'Agata, M. N., Szymanski, K. M., Ciecko, E. C., & Witman, M. A. (2026). Characterizing short‐term linear and nonlinear heart rate variability in typically developing children and adolescents. Physiological Reports, 14(19), Article e71126. https://doi.org/10.14814/phy2.71126

Image Credits: AI Generated

DOI: 10.14814/phy2.71126

Keywords: heart rate variability, autonomic nervous system, pediatrics, nonlinear dynamics, approximate entropy, detrended fluctuation analysis, Poincaré plot, puberty, parasympathetic tone, fractal dynamics, cardiac development, sex differences