Fully automated decoding of human activities and intentions from direct neural recordings is a tantalizing challenge in brain-computer interfacing. Implementing Brain Computer Interfaces (BCIs) outside carefully controlled experiments in laboratory settings requires adaptive and scalable strategies with minimal supervision. Here we describe an unsupervised approach to decoding neural states from naturalistic human brain recordings. We analyzed continuous, long-term electrocorticography (ECoG) data recorded over many days from the brain of subjects in a hospital room, with simultaneous audio and video recordings. We discovered coherent clusters in high-dimensional ECoG recordings using hierarchical clustering and automatically annotated them using speech and movement labels extracted from audio and video. To our knowledge, this represents the first time techniques from computer vision and speech processing have been used for natural ECoG decoding. Interpretable behaviors were decoded from ECoG data, including moving, speaking and resting; the results were assessed by comparison with manual annotation. Discovered clusters were projected back onto the brain revealing features consistent with known functional areas, opening the door to automated functional brain mapping in natural settings.

Motor behaviors are central to many functions and dysfunctions of the brain, and understanding their neural basis has consequently been a major focus in neuroscience. However, most studies of motor behaviors have been restricted to artificial, repetitive paradigms, far removed from natural movements performed "in the wild." Here, we leveraged recent advances in machine learning and computer vision to analyze intracranial recordings from 12 human subjects during thousands of spontaneous arm reach movements, observed over several days for each subject. These naturalistic movements elicited cortical spectral power patterns consistent with findings from controlled paradigms, but with an important difference: there was considerable neural variability across subjects and events. We modeled inter-event variability using ten behavioral and environmental features; the most important features explaining this variability were reach angle and recording day. Our work is among the first studies connecting behavioral and neural variability across cortex in humans during spontaneous movements and contributes to our understanding of long-term naturalistic behavior.

Recent evidence suggests that the aromatization of testosterone to estrogen is important for the organizing effects of neonatal testosterone on neuroendocrine responses to acute challenges. However, the extent to which neonatal inhibition of aromatase alters the stress-induced activation of neural pathways has not been examined. Here we assessed central patterns of c-fos mRNA induced by 30 min of restraint in 65-d-old adult male rats that were implanted with sc capsules of the aromatase inhibitor 1,4,6-androstatriene-3,17-dione (ATD), introduced within 12 h of birth and removed on d 21 of weaning. Neonatal ATD decreased the expression of arginine vasopressin within extrahypothalamic regions in adults, confirming reduced estrogen exposure during development. As adults, ATD-treated animals showed higher corticosterone responses at 30 min of restraint exposure compared with control animals as well as higher c-fos expression levels in the paraventricular nucleus of the hypothalamus. ATD treatment also increased stress-induced c-fos within several limbic regions of the forebrain, in addition to areas involved in somatosensory processing. Based on these results, we propose that the conversion of testosterone to estrogen during the neonatal period exerts marked, system-wide effects to organize adult neuroendocrine responses to homeostatic threat.

for discussions and help with data collection and analysis; neurosurgeons Jeffrey G. Ojemann and Andrew Ko, as well as the excellent staff at Harborview Hospital Neurosurgery department, for their care of the patients during their monitoring.