Mind Reading  

A newly developed neural network method now makes it easier and more accurate to decode fMRI scans. Using deep learning, researchers were able to reconstruct images a viewer was looking at through brain scan data analysis. The 'deep generative multiview model' also learned to correlate the data so that the accompanying standard fMRI noise could be accounted for in the generation of the reconstructed images.

One of the far reaching goals of neuroscience is to be able to read a person's thoughts. Such a technology is the inspiration, in part for Elon Musk's new company, Neuralink, along with other brain-machine interface ventures. So far, for data coming from functional magnetic resonance imaging (fMRI) scans, the task has proven to be very challenging.

fMRI scans are inherently noisy, and the activity in one voxel is well known to be influenced by activity in other voxels. This kind of correlation is computationally difficult and expensive to manage. Most work in this area has simply not dealt with it. This has significantly reduces the quality of the image reconstructions they produce.

Now, Changde Du at the Research Center for Brain-Inspired Intelligence in Beijing, China, and they his research team have developed a better ways to process data from fMRI scans to produce more accurate brain-image reconstructions. The team's research has been published online.

Their method uses deep learning techniques that handle nonlinear correlations between voxels more capably. The result is a much better way to reconstruct the way a brain perceives images.

Changde used several data sets of fMRI scans of the visual cortex of a human subject looking at a simple image—a single digit or a single letter. Each data set consists of the scans and the original image. They mapped the data to find a way to use the fMRI scans to reproduce the viewer's perceived image. In total, the team has access to over 1,800 fMRI scans and original images.

Related articles

• Google's Inceptionism Lets Us Look at an Artificial Intelligence Hallucination 
• Stanford Researchers and Google Create World's Largest Artificial Neural Network 
• The US Army Tries Mind Reading 

According to the researchers it was a straightforward deep learning task. They used 90 percent of the data to train the network to understand the correlation between the brain scan and the original image. Next, they tested the neural network on the remaining data by feeding it the scans and asking it to reconstruct what the viewed images were.

This approach had the advantage of having the network learn which voxels were used to reconstruct the image, avoiding the need to process the data from them all.

The neural network also learned how the data from the fMRI data was correlated. This was an important part of the research, because if the correlations are ignored, they end up being treated like noise and discarded. So the new approach—the so-called deep generative multiview model or DGMM—exploits these correlations and distinguishes them from real noise.  

The team compared their results from those of a number of other brain image reconstruction techniques. (See image at top). Generally, the reconstructed images are clear representations of the originals, and were for the most part superior to those derived by other methods.

“Extensive experimental comparisons demonstrate that our approach can reconstruct visual images from fMRI measurements more accurately,” write the study authors.

The research may have other implications other than regenerating what a view sees by interpreting a brain scan. "Although we focused on visual image reconstruction problem in this paper, our framework can also deal with brain encoding tasks," write the study authors.

The next steps for the research will include ways to analyze scenes more complex than simple numbered text and possibly moving images.

SOURCE  MIT Technology Review

By  33rd Square	Embed

Aging  

A new study shows that the difference between older brains and younger ones may not be so great. The researchers demonstrated that functional magnetic resonance imaging (fMRI), which is commonly used to study brain activity, is susceptible to signal noise from changing vascular activity.

Older brains may be more similar to younger brains than previously thought. In a new paper published in Human Brain Mapping, BBSRC-funded researchers at the University of Cambridge and Medical Research Council's Cognition and Brain Sciences Unit demonstrate that previously reported changes in the ageing brain using functional magnetic resonance imaging (fMRI) may be due to vascular (or blood vessels) changes, rather than changes in neuronal activity itself. Given the large number of fMRI studies used to assess the aging brain, this has important consequences for understanding how the brain changes with age and challenges current theories of ageing.

Related articles The Brain's Visual Perception Areas Mapped In Breakthrough Study  High Resolution Images Show Brain Activity Like Never Before  First Images Of Infant Brain Connectivity In The Womb Made

A fundamental problem of fMRI is that it measures neural activity indirectly through changes in regional blood flow. Thus, without careful correction for age differences in vasculature reactivity, differences in fMRI signals can be erroneously regarded as neuronal differences. An important line of research focuses on controlling for noise in fMRI signals using additional baseline measures of vascular function. However, such methods have not been widely used, possibly because they are impractical to implement in studies of ageing.

"These findings clearly show that without such correction methods, fMRI studies of the effects of age on cognition may misinterpret effect of age as a cognitive, rather than vascular, phenomena."

An alternative candidate for correction makes use of resting state fMRI measurements, which is easy to acquire in most fMRI experiments. While this method has been difficult to validate in the past, the unique combination of an impressive data set across 335 healthy volunteers over the lifespan, as part of the CamCAN project, allowed Dr. Kamen Tsvetanov and colleagues to probe the true nature of ageing effects on resting state fMRI signal amplitude. Their research showed that age differences in signal amplitude during a task are of a vascular, not neuronal, origin. They propose that their method can be used as a robust correction factor to control for vascular differences in fMRI studies of ageing.

The study also challenged previous demonstrations of reduced brain activity in visual and auditory areas during simple sensorimotor tasks. Using conventional methods, the current study replicated these findings. However, after correction, Tsvetanov and his team's results show that it might be vascular health, not brain function, that accounts for most age-related differences in fMRI signal in sensory areas. Their results suggest that the age differences in brain activity may be overestimated in previous fMRI studies of ageing.

Tsvetanov said: "There is a need to refine the practice of conducting fMRI. Importantly, this doesn't mean that studies lacking 'golden standard' calibration measures, such as large scale studies, patient studies or ongoing longitudinal studies are invalid. Instead, researchers should make use of available resting state data as a suitable alternative. These findings clearly show that without such correction methods, fMRI studies of the effects of age on cognition may misinterpret effect of age as a cognitive, rather than vascular, phenomena."


SOURCE  BBSRC

By 33rd Square

Embed

 Neuroscience  

New research suggests that the various reports—of both over-and under-connectivity—may, in fact, reflect a deeper principle of brain function. The study shows that the brains of individuals with autism display unique synchronization patterns, something that could impact earlier diagnosis of the disorder and future treatments.

New research recently published in Nature Neuroscience suggests that the various reports—of both over-and under-connectivity—may, in fact, reflect a deeper principle of brain function. Led by scientists at the Weizmann Institute and Carnegie Mellon University, the study shows that the brains of individuals with autism display unique synchronization patterns, something that could impact earlier diagnosis of the disorder and future treatments.

"Identifying brain profiles that differ from the pattern observed in typically developing individuals is crucial not only in that it allows researchers to begin to understand the differences that arise in ASD but, in this case, it opens up the possibility that there are many altered brain profiles all of which fall under the umbrella of 'autism' or 'autisms,'" said Marlene Behrmann, the George A. and Helen Dunham Cowan Professor of Cognitive Neuroscience at Carnegie Mellon and co-director of the Center for the Neural Basis of Cognition.

"Identifying brain profiles that differ from the pattern observed in typically developing individuals is crucial not only in that it allows researchers to begin to understand the differences that arise in ASD."

To investigate the issue of connectivity in ASD, the researchers analyzed data obtained from functional magnetic resonance imaging (fMRI) studies conducted while the participants were at rest. Data was collected from a large number of participants at multiple sites and handily assembled in the ABIDE database.

"Resting-state brain studies are important because that is when patterns emerge spontaneously, allowing us to see how various brain areas naturally connect and synchronize their activity," said Avital Hahamy, a Ph.D. student in Weizmann's Neurobiology Department.

Related articles Carnegie Mellon's HERB Masters Critical Oreo Separation Task  The Brain's Visual Perception Areas Mapped In Breakthrough Study  Potential Cause of Autism Discovered

A number of previous studies by these researchers and others suggest that these spontaneous patterns may provide a window into individual behavioral traits, including those that stray from the norm.
In a careful comparison of the details of these intricate synchronization patterns, the scientists discovered an intriguing difference between the control and ASD groups: the control participants' brains had substantially similar connectivity profiles across different individuals, while those with ASD showed a remarkably different phenomenon. Those with autism tended to display much more unique patterns—each in its own, individual way. They realized that the synchronization patterns seen in the control group were "conformist" relative to those in the ASD group, which they termed "idiosyncratic."

Differences between the synchronization patterns in the autism and control groups could be explained by the way individuals in the two groups interact and communicate with their environment.

"From a young age, the average, typical person's brain networks get molded by intensive interaction with people and the mutual environmental factors," Hahamy said. "Such shared experiences could tend to make the synchronization patterns in the control group's resting brains more similar to each other. It is possible that in ASD, as interactions with the environment are disrupted, each one develops a more uniquely individualistic brain organization pattern."

The researchers emphasize that this explanation is only tentative; much more research will be needed to fully uncover the range of factors that may lead to ASD-related idiosyncrasies. They also suggest that further research into how and when different individuals establish particular brain patterns could help in the future development of early diagnosis and treatment for autism disorders.


SOURCE  Carnegie Mellon University via EurekAlert

By 33rd Square

Embed

 Neuroscience  

Noninvasive brain scans have led to basic science discoveries about the human brain, but they've had only limited impacts on people's day-to-day lives. A number of recent studies showing that brain imaging can help predict an individual's future learning, criminality, health-related behaviors, and response to drug or behavioral treatments.

Noninvasive brain scans, such as functional magnetic resonance imaging (FMRI), have led to basic science discoveries about the human brain, but they've had only limited impacts on people's day-to-day lives. A review article published in the journal Neuron, however, highlights a number of recent studies showing that brain imaging can help predict an individual's future learning, criminality, health-related behaviors, and response to drug or behavioral treatments. The technology may offer opportunities to personalize educational and clinical practices.

"We often wait for failure, in school or in mental health, to prompt attempts to help, but by then a lot of harm has occurred. If we can use neuroimaging to identify individuals at high risk for future failure, we may be able to help those individuals avoid such failure altogether."

Dr. John Gabrieli of the Massachusetts Institute of Technology and his colleagues describe the predictive power of brain imaging across a variety of different future behaviors, including infants' later performance in reading, students' later performance in math, criminals' likelihood of becoming repeat offenders, adolescents' future drug and alcohol use, and addicts' likelihood of relapse.

"Presently, we often wait for failure, in school or in mental health, to prompt attempts to help, but by then a lot of harm has occurred," says Dr. Gabrieli. "If we can use neuroimaging to identify individuals at high risk for future failure, we may be able to help those individuals avoid such failure altogether."

Prior to treatment, patients with social anxiety disorder who exhibited greater posterior activation (left) for angry relative to neutral facial expressions had a better clinical response to cognitive behavioral therapy than patients who exhibited lesser activation (right)

Related articles Dennis Hong On Robot Evolution and The DARPA Robotics Challenge  Boston Dynamic's Big Dog Gets An Arm For A Head  The What, Why, How, Who and When of the DARPA Robotics Challenge

In particular, Gabrieli focuses on education as he believes brain imaging might help teachers identify those kids who will struggle when learning to read or do math. “Current behavioral testing is pretty good at identifying which children are at potential risk, but it’s still too hit-and-miss to trigger serious help (about half of children who look to be at potential risk turn out not to be at true risk),” he told Medical Daily. “Also, I can imagine that more knowledge from brain-informed outcomes might lead to new kinds of behavioral testing that could be more readily used in schools.”

The study authors also point to the clear ethical and societal issues that are raised by studies attempting to predict individuals' behavior. "Because of their biological nature, brain measures can be overly valued and potentially divert public and scientific interest in behavioral and social factors," they write.

"We will need to make sure that knowledge of future behavior is used to personalize educational and medical practices, and not be used to limit support for individuals at higher risk of failure," says Dr. Gabrieli. "For example, rather than simply identifying individuals to be more or less likely to succeed in a program of education, such information could be used to promote differentiated education for those less likely to succeed with the standard education program."

SOURCE  Cell Press via EurekAlert

By 33rd Square

Embed

 Consciousness  

Psychologists have used brain-imaging techniques to study what happens to the human brain when it slips into unconsciousness. Their research is an initial step toward developing a scientific definition of consciousness.

Psychologists have used brain-imaging techniques to study what happens to the human brain when it slips into unconsciousness. Their research, published in the online journal PLOS Computational Biology, is an initial step toward developing a scientific definition of consciousness.

"In terms of brain function, the difference between being conscious and unconscious is a bit like the difference between driving from Los Angeles to New York in a straight line versus having to cover the same route hopping on and off several buses that force you to take a 'zig-zag' route and stop in several places," said lead study author Martin Monti, an assistant professor of psychology and neurosurgery at UCLA.

Monti and his colleagues used functional magnetic resonance imaging (fMRI) to study how the flow of information in the brains of 12 healthy volunteers changed as they lost consciousness under anesthesia with propofol. The participants ranged in age from 18 to 31 and were evenly divided between men and women.

Image Source: Monti et al, PLOS Computational Biology

The psychologists analyzed the "network properties" of the subjects' brains using a branch of mathematics known as graph theory, which is often used to study air-traffic patterns, information on the Internet and social groups, among other topics.

"It turns out that when we lose consciousness, the communication among areas of the brain becomes extremely inefficient, as if suddenly each area of the brain became very distant from every other, making it difficult for information to travel from one place to another," Monti said.

The finding shows that consciousness does not "live" in a particular place in our brain but rather "arises from the mode in which billions of neurons communicate with one another," he said.

Related articles US Set To Enter The Brain Race With The Brain Activity Map Project  The Brain's Visual Perception Areas Mapped In Breakthrough Study  Consciousness: The Hard Problem

When patients suffer severe brain damage and enter a coma or a vegetative state, Monti said, it is very possible that the sustained damage impairs their normal brain function and the emergence of consciousness in the same manner as was seen by the life scientists in the healthy volunteers under anesthesia.

"If this were indeed the case, we could imagine in the future using our technique to monitor whether interventions are helping patients recover consciousness," he said.

"It could, however, also be the case that losing consciousness because of brain injury affects brain function through different mechanisms," said Monti, whose research team is currently addressing this question in another study.

"As profoundly defining of our mind as consciousness is, without having a scientific definition of this phenomenon, it is extremely difficult to study," Monti noted. This study, he said, marks an initial step toward conducting neuroscience research on consciousness.

The research was conducted at Belgium's University Hospital of Liege.

Monti's expertise includes cognitive neuroscience, the relationship between language and thought, and how consciousness is lost and recovered after severe brain injury. He was part of a team of American and Israeli brain scientists who used fMRI on former Israeli Prime Minister Ariel Sharon in January 2013 to assess his brain responses.

Surprisingly, Sharon, who was presumed to be in a vegetative state since suffering a brain hemorrhage in 2006, showed significant brain activity, Monti and his colleagues reported.

The former prime minister was scanned to assess the extent and quality of his brain processing, using methods recently developed by Monti and his colleagues. The scientists found subtle but encouraging signs of consciousness.



SOURCE  UCLA

By 33rd Square

Subscribe to 33rd Square

 Neuroimaging  

Using a high ​​field strength fMRI magnet, researchers have achieved images of the human brain stem at resolutions not previously possible. The work has let them peer into the tiny PAG gray region, an area not previously seen in such detail.

D eep in our brains  there is a tiny struc­ture shaped like an elon­gated donut that plays a cru­cial role in man­aging how the body func­tions. Mea­suring just 10 mil­lime­ters in length and six mil­lime­ters in diam­eter, the hollow struc­ture is involved in a com­plex array of behav­ioral, cog­ni­tive, and affec­tive phe­nomena, such as the fight or flight response, pain reg­u­la­tion, and even sexual activity, according to senior research sci­en­tist Ajay Satpute.

With a name longer than the struc­ture itself, the “mid­brain peri­aque­ductal gray region,” or PAG, is extra­or­di­narily dif­fi­cult to inves­ti­gate in humans because of its size and intri­cate struc­ture, he said.

Now, in research published online in the journal Pro­ceed­ings of the National Academy of Sci­ence, Sat­pute and his col­leagues at Northeastern University's Inter­dis­ci­pli­nary Affec­tive Sci­ence Lab­o­ra­tory explain how they overcame thes challenges using  state-​​of-​​the art imaging to cap­ture this com­plex neural activity.

The research could ultimately help scientists explore the grounds of human emotion like never before.

“The PAG’s func­tional prop­er­ties occur at such small spa­tial scales that we need to cap­ture its activity at very high res­o­lu­tion in order to under­stand it,” he explained.

Isolation of the PAG - Image Source: Ajay Satpute et. al.

Related articles Mini Human Brain Grown In Petri Dish  New Technique Allows Individual Synapses To Be Mapped  First Images Of Infant Brain Connectivity In The Womb Made

Until recently, neu­roimaging studies have been car­ried out on func­tional mag­netic res­o­nance imaging, or fMRI, instruments containing mag­nets of up to three Teslas, a mea­sure of mag­netic field strength. These instru­ments pro­vide crit­ical data for under­standing how the brain’s dif­ferent areas respond to dif­ferent stimuli, but when those areas become suf­fi­ciently small and com­pli­cated, their res­o­lu­tion falls short.
In the case of the tiny PAG, this problem is para­mount because the PAG wraps around a hollow core, or “aque­duct,” con­taining cere­brospinal fluid, Sat­pute said.

Tra­di­tional fMRI instru­ments cannot dis­tin­guish neural activity occur­ring in the PAG from that occur­ring in the CS fluid. Even more dif­fi­cult is iden­ti­fying where within the PAG itself spe­cific responses originate.

In col­lab­o­ra­tion with researchers at the Mass­a­chu­setts Gen­eral Hos­pital in Boston, Sat­pute and his col­leagues used a high-​​tech fMRI instru­ment that con­tains a seven-​​Tesla magnet. The force of the instru­ment is so strong (albeit harm­less) that one can feel its pull when simply walking by. Cou­pled with painstaking manual data analyses, Sat­pute was able to resolve activity in sub-​​regions of the PAG with more pre­ci­sion than ever before.

With their method in hand, the research team showed 11 human research sub­jects images of burn vic­tims, gory injuries, and other con­tent related to threat, harm, and loss while keeping tabs on the PAG’s activity. Researchers also showed the sub­jects neu­tral images such and then com­pared results between the two scenarios.

The proof-​​of-​​concept study showed emotion-​​related activity con­cen­trated in par­tic­ular areas of the PAG. While sim­ilar results have been demon­strated in animal models, nothing like it had pre­vi­ously been shown in human brains.

Using this method­ology, the researchers said they would not only gain a better under­standing of the PAG but also be able to inves­ti­gate a range of brain-​​related research ques­tions beyond this par­tic­ular structure.

Seven-​​Tesla brain imaging pro­vides an unprece­dented view of regions like the PAG while they respond to stimuli, said Lisa Feldman Bar­rett, director of the Inter­dis­ci­pli­nary Affec­tive Sci­ence Lab­o­ra­tory. “Studies like this are a crit­ical step for­ward in bridging human and non­human animal studies of emo­tion, because they offer a level of res­o­lu­tion in human brains that was pre­vi­ously pos­sible only in studies of non-​​human animal,” she said.



SOURCE  Northeastern University

By 33rd Square

Subscribe to 33rd Square

 Mind Reading  

Using a mathematical model, researchers in The Netherlands have reconstructed thoughts from data collected from fMRI test subjects - essentially they read the minds of the participants.

By analysing MRI images of the brain with an elegant mathematical model, researchers from Radboud University Nijmegen have reconstruct thoughts more accurately than ever before. In this way, they have succeeded in determining which letter a test subject was looking at.

The researchers work has been published in the  journal Neuroimage.

Functional MRI scanners have been used in cognition research primarily to determine which brain areas are active while test subjects perform a specific task. The question is simple: is a particular brain region on or off? A research group at the Donders Institute for Brain, Cognition and Behaviour at Radboud University has gone a step further: they have used data from the scanner to determine what a test subject is looking at.

The researchers 'taught' a model how small volumes of 2x2x2 mm from the brain scans -- known as voxels -- respond to individual pixels. By combining all the information about the pixels from the voxels, it became possible to reconstruct the image viewed by the subject. The result was not a clear image, but a somewhat fuzzy speckle pattern. In this study, the researchers used hand-written letters.

Related articles Breakthrough Technique Gives Unprecedented View Inside The Brain  BigBrain Presents An Ultrahigh-Resolution 3D Model of the Human Brain  Researchers Demonstrate Matrix-Style Learning

"After this we did something new", says lead researcher Marcel van Gerven. "We gave the model prior knowledge: we taught it what letters look like. This improved the recognition of the letters enormously. The model compares the letters to determine which one corresponds most exactly with the speckle image, and then pushes the results of the image towards that letter. The result was the actual letter, a true reconstruction."

"Our approach is similar to how we believe the brain itself combines prior knowledge with sensory information. For example, you can recognize the lines and curves in this article as letters only after you have learned to read. And this is exactly what we are looking for: models that show what is happening in the brain in a realistic fashion. We hope to improve the models to such an extent that we can also apply them to the working memory or to subjective experiences such as dreams or visualisations. Reconstructions indicate whether the model you have created approaches reality."

In other words, the researchers claim to be very close to the ability to read your mind with their technique.  Such an understanding may also open up the possibility of implanting thoughts, knowledge or, on a more sinister level, control the actions of individuals without their authority.

"In our further research we will be working with a more powerful MRI scanner," explains Sanne Schoenmakers, who is working on a thesis about decoding thoughts. "Due to the higher resolution of the scanner, we hope to be able to link the model to more detailed images. We are currently linking images of letters to 1200 voxels in the brain; with the more powerful scanner we will link images of faces to 15,000 voxels."


SOURCE  Radboud University Nijmegen

By 33rd Square

Subscribe to 33rd Square