Miguel A. L. Nicolelis

In recent years, our group and others have reported multiple cases of consistent neurological recovery in people with spinal cord injury (SCI) following a protocol that integrates locomotion training with brain machine interfaces (BMI). The primary objective of this pilot study was to compare the neurological outcomes (motor, tactile, nociception, proprioception, and vibration) in both an intensive assisted locomotion training (LOC) and a neurorehabilitation protocol integrating assisted locomotion with a noninvasive brain–machine interface (L + BMI), virtual reality, and tactile feedback. We also investigated whether individuals with chronic-complete SCI could learn to perform leg motor imagery. We ran a parallel two-arm randomized pilot study; the experiments took place in São Paulo, Brazil. Eight adults sensorimotor-complete (AIS A) (all male) with chronic (> 6 months) traumatic spinal SCI participated in the protocol that was organized in two blocks of 14 weeks of training and an 8-week follow-up. The participants were allocated to either the LOC group (n = 4) or L + BMI group (n = 4) using block randomization (blinded outcome assessment). We show three important results: (i) locomotion training alone can induce some level of neurological recovery in sensorimotor-complete SCI, and (ii) the recovery rate is enhanced when such locomotion training is associated with BMI and tactile feedback (∆Mean Lower Extremity Motor score improvement for LOC = + 2.5, L + B = + 3.5; ∆Pinprick score: LOC = + 3.75, L + B = + 4.75 and ∆Tactile score LOC = + 4.75, L + B = + 9.5). (iii) Furthermore, we report that the BMI classifier accuracy was significantly above the chance level for all participants in L + B group. Our study shows potential for sensory and motor improvement in individuals with chronic complete SCI following a protocol with BMIs and locomotion therapy. We report no dropouts nor adverse events in both subgroups participating in the study, opening the possibility for a more definitive clinical trial with a larger cohort of people with SCI.Trial registration: http://www.ensaiosclinicos.gov.br/ identifier RBR-2pb8gq.

Back in 2011, my laboratory at the Duke University Center for Neuroengineering proposed a new variation on the already-classic paradigm known as brain–machine interfaces (BMIs) [ 1]. In a series of experiments [ 2], we showed that monkeys could not only learn to employ electrical signals from large neuronal ensembles (300–700 neurons) located in the primary motor cortex (M1) to control the movements of an avatar arm and hand—i.e. using a typical BMI paradigm—but that they could also interpret in real time, streams of electrical signals delivered directly to their primary somatosensory cortex (S1) via cortical microstimulation. This strategy was created to allow these monkeys to perform a tactile discrimination task, without any interference from their physical bodies. Thus, instead of using their real finger tips to perform such a task, this approach allowed our monkeys to use their brain activity to directly control the movements of ‘virtual finger tips’ to scan the texture of a series of virtual objects. Upon receiving, via cortical microstimulation, electrical signals directly into their S1 that were proportional to the virtual surface textures of these objects, the monkeys learned to quickly select the correct object that they needed to identify to receive a reward. Essentially, in this pioneering study and in a series of other experiments that followed it in the operation of what we named brain–machine–brain interfaces (BMBIs) (Fig. 1) [ 3, 4], we demonstrated and introduced a paradigm that a decade later has the potential to guide a true revolution in the way we design and implement a new generation of neuroprostheses. Indeed, I envision that new technologies that at their core employ the BMBI concept may lead to a series of novel non-pharmacological therapies aimed at treating a series of neurological and psychiatric disorders that afflict hundreds of millions of people worldwide—the therapeutic management of which, even today, remains very challenging.

Although international airports served as main entry points for SARS-CoV-2, the factors driving the uneven geographic spread of COVID-19 cases and deaths in Brazil remain mostly unknown. Here we show that three major factors influenced the early macro-geographical dynamics of COVID-19 in Brazil. Mathematical modeling revealed that the “super-spreading city” of São Paulo initially accounted for more than 85% of the case spread in the entire country. By adding only 16 other spreading cities, we accounted for 98–99% of the cases reported during the first 3 months of the pandemic in Brazil. Moreover, 26 federal highways accounted for about 30% of SARS-CoV-2’s case spread. As cases increased in the Brazilian interior, the distribution of COVID-19 deaths began to correlate with the allocation of the country’s intensive care units (ICUs), which is heavily weighted towards state capitals. Thus, severely ill patients living in the countryside had to be transported to state capitals to access ICU beds, creating a “boomerang effect” that contributed to skew the distribution of COVID-19 deaths. Therefore, if (i) a lockdown had been imposed earlier on in spreader-capitals, (ii) mandatory road traffic restrictions had been enforced, and (iii) a more equitable geographic distribution of ICU beds existed, the impact of COVID-19 in Brazil would be significantly lower.

Advances in neuroscience are inspiring developments in robotics and vice versa.

Although its international airports served as the country’s main entry points for SARS-CoV-2, the factors driving the uneven geographic spread of COVID-19 cases and deaths in Brazil remain largely unknown. Here we show that four major factors likely accounted for the entire dynamics of COVID-19 in Brazil. Mathematical modeling revealed that, initially, the “super-spreading city” of São Paulo accounted for roughly 80% of the case spread in the entire country. During the first 3 months of the epidemic, by adding only 16 other spreading cities, we accounted for 98-99% of the cases reported in Brazil at the time. Moreover, 26 of the major Brazilian federal highways accounted for about 30% of SARS-CoV-2’s case spread. As cases accumulated rapidly in the Brazilian countryside, the distribution of COVID-19 deaths began to correlate with a third parameter: the geographic distribution of the country’s hospital intensive care unit (ICU) beds, which is highly skewed towards state capitals where the epidemic began. That meant that severely ill patients living in the countryside had to be transported to state capitals to access ICU beds where they often died, creating a “boomerang effect” that contributed to the skew of the geographic distribution of COVID-19 deaths. Finally, we discovered that the geographic distribution of dengue fever, amounting to more than 3.5 million cases from January 2019 to July 2020, was highly complementary to that of COVID-19. This was confirmed by the identification of significant negative correlations between COVID-19’s incidence, infection growth rate, and mortality to the percentage of people with antibody (IgM) levels for dengue fever in each of the country’s states. No such correlations were observed when IgM data for chikungunya virus, which is transmitted by the same mosquito vector as dengue, was used. Thus, states in which a large fraction of the population had contracted dengue fever in 2019-2020 reported lower COVID-19 cases and deaths, and took longer to reach exponential community transmission, due to slower SARS-CoV-2 infection growth rates. This inverse correlation between COVID-19 and dengue fever was further observed in a sample of countries around Asia and Latin America, as well as in islands in the Pacific and Indian Oceans. This striking finding raises the intriguing possibility of an immunological cross-reactivity between DENV serotypes and SARS-CoV-2. If proven correct, this hypothesis could mean that dengue infection or immunization with an efficacious and safe dengue vaccine could produce some level of immunological protection for SARS-CoV-2, before a vaccine for SARS-CoV-2 becomes available.

Lack of sensory feedback is a major obstacle in the rapid absorption of prosthetic devices by the brain. While electrical stimulation of cortical and subcortical structures provides unique means to deliver sensory information to higher brain structures, these approaches require highly invasive surgery and are dependent on accurate targeting of brain structures. Here, we propose a semi-invasive method, Dorsal Column Stimulation (DCS) as a tool for transferring sensory information to the brain. Using this new approach, we show that rats can learn to discriminate artificial sensations generated by DCS and that DCS-induced learning results in corticostriatal plasticity. We also demonstrate a proof of concept brain-to-spine interface (BTSI), whereby tactile and artificial sensory information are decoded from the brain of an “encoder” rat, transformed into DCS pulses, and delivered to the spinal cord of a second “decoder” rat while the latter performs an analog-to-digital conversion during a sensory discrimination task. These results suggest that DCS can be used as an effective sensory channel to transmit prosthetic information to the brain or between brains, and could be developed as a novel platform for delivering tactile and proprioceptive feedback in clinical applications of brain-machine interfaces.

Back in 2012, Churchland and his colleagues proposed that “rotational dynamics”, uncovered through linear transformations of multidimensional neuronal data, represent a fundamental type of neuronal population processing in a variety of organisms, from the isolated leech central nervous system to the primate motor cortex. Here, we evaluated this claim using Churchland’s own data and simple simulations of neuronal responses. We observed that rotational patterns occurred in neuronal populations when (1) there was a temporal sequence in peak firing rates exhibited by individual neurons, and (2) this sequence remained consistent across different experimental conditions. Provided that such a temporal order of peak firing rates existed, rotational patterns could be easily obtained using a rather arbitrary computer simulation of neural activity; modeling of any realistic properties of motor cortical responses was not needed. Additionally, arbitrary traces, such as Lissajous curves, could be easily obtained from Churchland’s data with multiple linear regression. While these observations suggest that temporal sequences of neuronal responses could be visualized as rotations with various methods, we express doubt about Churchlandet al.’s bold assessment that such rotations are related to “an unexpected yet surprisingly simple structure in the population response”, which “explains many of the confusing features of individual neural responses”. Instead, we argue that their approach provides little, if any, insight on the underlying neuronal mechanisms employed by neuronal ensembles to encode motor behaviors in any species.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Lack of sensory feedback is a major obstacle in the rapid absorption of prosthetic devices by the brain. While electrical stimulation of cortical and subcortical structures provides unique means to deliver sensory information to higher brain structures, these approaches require highly invasive surgery and are dependent on accurate targeting of brain structures. Here, we propose a semi-invasive method, Dorsal Column Stimulation (DCS) as a tool for transferring sensory information to the brain. Using this new approach, we show that rats can learn to discriminate artificial sensations generated by DCS and that DCS-induced learning results in corticostriatal plasticity. We also demonstrate a proof of concept brain-to-spine interface (BTSI), whereby tactile and artificial sensory information are decoded from the brain of an “encoder” rat, transformed into DCS pulses, and delivered to the spinal cord of a second “decoder” rat while the latter performs an analog-to-digital conversion during a tactile discrimination task. These results suggest that DCS can be used as an effective sensory channel to transmit prosthetic information to the brain or between brains, and could be developed as a novel platform for delivering tactile and proprioceptive feedback in clinical applications of brain-machine interfaces.