Emery N. Brown

Brown developed statistical methods to characterize the properties of the human circadian system (biological clock) from core temperature data recorded under the constant routine and free-running and forced desynchrony protocols. Through the early part of his career, Brown collaborated with circadian researchers to apply his methods to answer fundamental research questions in circadian physiology. Brown's statistical methods were critical for: estimating accurately the period and internal time on human circadian clocks from continuous core temperature measurement;[10][11] showing that bright lights could be used to shift the phase of the human circadian clock;[12] properly timed administration of light and dark periods could be used to realign the internal clocks of shift workers with external time;[13] and that, contrary to beliefs at the time, the period of the human biological clock, like that of other animals, was closer to 24 hours rather than 25 hours.[14]

Brown later focused his statistics research on developing signal processing algorithms and statistical methods for neuronal data analysis. He developed a state-space point process (SSPP) paradigm to study how neural systems maintain dynamic representations of information.[15] For the analysis of neural spiking activity and binary behavioral tasks represented as multivariate or univariate point processes (0-1 events that occur in continuous time), his research produced analogs of the Kalman filter, Kalman smoothing, sequential Monte Carlo algorithms, and combined state and parameter estimation algorithms commonly applied to continuous-valued time series observations.

Brown used the methods to: show that ensembles of neurons in the rodent hippocampus maintained a highly accurate representation of the animal's spatial location;[16] track the formation of neural receptive fields on a millisecond time scale;[17][18][19] track concurrent changes in neural activity and behavior during learning experiments;[20] decode how groups of motor neurons represent movement information;[21] and track burst suppression in patients under general anesthesia.[22]

Brown applied the state-space paradigm to: analyze learning in behavioral neuroscience experiments;[23][24][25] study the relationship between learning and changes in hippocampal function in humans;[26] assess the efficacy of deep brain stimulation in enhancing behavior performance in humans and non-human primates;[27] and define precisely changes in levels of consciousness under propofol-induced general anesthesia.[28]

With Partha Mitra, Brown co-founded and co-directed the Neuroinformatics Summer Course at the Marine Biological Laboratory in Woods Hole, MA from 2002 to 2006. He co-directs with Robert Kass the biannual Statistical Analysis of Neural Data Conference at the Carnegie Mellon University Center for the Neural Basis of Cognition.[29][30] He co-authored a textbook in neuroscience data analysis with Robert Kass and Uri Eden.[31]

Unraveling the mystery of general anesthesia is another major question facing modern medicine.[5] In 2004, Brown began a systems neuroscience research program to study the mechanisms of anesthetic action by forming and leading an interdisciplinary collaboration of anesthesiologists, neuroscientists, a statistician, a neurosurgeon, neurologists, bioengineers and a mathematician at MGH, MIT and Boston University.[32] In 2007 he received an NIH Director's Pioneer Award to support this research making him, the first anesthesiologist and the first statistician to receive this award.[33] His anesthesiology research has made fundamental theoretical and experimental contributions to understanding the neurophysiology of general anesthesia. In two seminal papers,[34][35] Brown provided the first systems neuroscience analysis of how anesthetics act at specific receptors in specific neural circuits to produce commonly observed altered arousal states. This analysis provided an essential missing link between the substantial body of research on the molecular pharmacology of anesthetic action and the behavioral responses commonly seen in anesthetized patients. Brown also shows that, contrary to common dogma general anesthesia is not sleep, but rather a reversible coma.[34]

Brown's research group has provided detailed insights into how anesthetics produce unconsciousness. The brain is not shut off under general anesthesia. Instead, anesthetics induce highly structured oscillations between key brain regions. These oscillations, which are readily visible in standard electroencephalogram (EEG) recordings, alter arousal by impairing normal communication between regions. This is analogous to what happens when an epilepsy patient loses consciousness with the appearance of the regular, hypersynchronous oscillations of a seizure. Anesthetic-induced oscillations are also akin to what happens when a hum in a phone line makes it impossible to sustain a normal conversation.[34][35]

Brown has performed many studies on the properties of propofol-induced anesthesia in particular. He found that propofol-induced unconsciousness is mediated simultaneously by two different oscillatory processes. The first is strong coherent alpha oscillations (8 to 10 cycles per second) between the cortex and the thalamus (26–28) and the second are strong incoherent cortical slow-wave oscillations (<1 cycle per second).[36][37][38] The alpha oscillations impair communication between the thalamus and cortex. The slow-waves restrict to narrow time intervals the times at which cortical neurons can discharge, thus making it difficult to sustain communication within the cortex.[37] Furthermore, each anesthetic has a different EEG signature reflecting different neural circuit mechanisms action. These signatures change with age and the anesthetic dose.[39][40] A practical implication of this finding is that the EEG can be used in real time to monitor accurately the anesthetic state of patients. Brown's group has developed an online teaching program to train anesthesiologists on this monitoring approach.[41]

Brown and colleagues are establishing a new paradigm for waking patients up following general anesthesia. They have shown that the anesthetic state can be rapidly reversed by administering methylphenidate (Ritalin)[42] or activation of dopaminergic systems.[43] This suggests a new, feasible way to actively restore cognitive function in patients after anesthesia and sedation. They have received FDA approval to undertake a clinical trial to test this idea in humans (NCT 02051452).[44][45][46] They have also shown that burst-suppression, a state of profound brain inactivation seen in deep general anesthesia, hypothermia, coma and developmental brain disorders, can be simply explained by a unifying neural-metabolic model.[47] Brown's group have also shown that burst suppression can be precisely controlled to maintain a therapeutic, medically induced coma. This research uses a closed-loop control system based on his SSPP paradigm.[48][49] This could have important implications for treating patients, such as Gabby Giffords, Michael Schumaker, Malala Yousafzai, and Joan Rivers, who sustain brain injuries or have intracranial hypertension and require a medically induced coma to facilitate brain recovery.

Brown's anesthesiology research has been featured on National Public Radio,[50] in Scientific American,[51] the MIT Technology Review,[5] the New York Times[52] and in TEDMED[53] 2014.