I had a chance to stop by a really intriguing poster this morning on Day 5 (the final day!) of Neuroscience 2013. Brian Mills, a co-author on a poster (765.01) entitled “Model-based functional brain connectivity,” talked me through the pretty complex methods of this project.

In general, functional connectivity studies involving resting state fMRI entail comparing connectivity patterns between different groups of participants (e.g., schizophrenia vs. typically developing). Instead, Mills and colleagues are exploring whether or not resting state fMRI can be used to identify connectivity patterns that are unique to a specific individual. They are seeking an fMRI fingerprint.

In their study, Mills and colleagues developed a model-based connectivity matrix for a set of regions of interest (ROIs) throughout the brain of a participant, in which each cell of the matrix represents not the connectivity between two ROIs (as is traditional in functional connectivity analyses), but instead the relative contribution of neighboring ROIs to the timecourse of a specific ROI. This matrix can be used to model and predict the connectivity patterns of a set of ROIs at another time, such as a second fMRI scan date.

Comparing different linear equations that produce these model matrices, Mills and colleagues have determined the best way to predict one’s resting state fMRI connectivity patterns. And by best, I mean pretty darn accurate. Using their modeling techniques, Mills and colleagues have been able to classify single participants at scans a week after their original scan with up to 100% accuracy.

Even better, this type of modeling only seems to require about 60 frames of resting-state fMRI data to achieve accurate classification of individual participants. These frames, totaling about 2.5 minutes of scanning, don’t even have to be in sequential order. This is especially useful for patient populations, such as autism, in which motion artifact is a particular concern. Mills noted that with this method you can use participants who may otherwise not be included in data analysis due to too much motion in the scanner. This classification method also works in monkeys, proving that human-specific behavior, such as mind-wandering, is not leading to the accuracy of this method.

Additionally, when looking across their pool of participants, Mills and colleagues were able to determine which brain regions were relatively stable, versus dynamic, in terms of connectivity patterns from individual to individual. It appears that connectivity in motor and sensory areas of the brain is conserved across individuals, whereas that of fronto-parietal areas and default mode network areas is more variable between individuals. These findings make natural sense, as I would think that the brain activity that makes someone unique would occur in more frontal, higher order networks of the brain, as opposed to more evolutionarily old regions, such as motor areas.
​
Mills noted that they hope to apply this method to explore how the classification of individuals changes across development and when patient populations such as ADHD are considered. Mills hopes that much like knowing the variability of one’s genetic makeup, this classification method will someday aid in predicting risk for different disorders as well as the likelihood of response to certain drugs.

​[This post was originally published at my previous blog, Neurolore.]

On Monday, I sat in on part of a nanosymposim entitled “Biomarkers and Imaging in Schizophrenia.” Again, I was drawn to these talks because of their relevance to my work, but their content is worth sharing with the broader SfN 2013 audience.
In the first talk I heard, Guusje Collin from the Rudolf Magnus Institute of Neuroscience presented data on the structural organization of networks in unaffected siblings of schizophrenia patients. Using diffusion tensor imaging (DTI), Collin and colleagues were able to look at the white matter tracts of the brains of these siblings in the form of a network. They were then able examine the connectivity within what’s called the “rich club” of these networks, meaning a group of densely interconnected hubs (pictured in red in the image) that serve as centers of  integration of information in the brain. The connectivity of these rich club regions has been previously demonstrated to be reduced in patients with schizophrenia.

In their current study, Collin have found that when looking at siblings of these individuals (who don’t have schizophrenia themselves), there is reduced rich club connectivity as compared to unrelated control participants. In fact, the unaffected siblings appear to have intermediate rich club connectivity that falls between the schizophrenia patients and the control participants. This finding indicates that there appears to be a familial or perhaps genetic influence on rich club connectivity as it relates to schizophrenia. Thus, this phenotype of impaired connectivity could indicate a predisposition to this disorder.

In future work, Collin plans to explore rich club connectivity in additional disorders, such as bipolar disorder, and compare the connectivity to that of schizophrenia.
In another talk, Hengyi Cao from Heidelberg University discussed an additional intermediate phenotype that can be found in unaffected siblings of schizophrenia patients. It is well documented that emotional dysfunction is a common symptom of schizophrenia. Previous studies involving functional connectivity analyses of unaffected siblings of schizophrenia patients have found no changes in connectivity in the amygdala, a region known to play a role in the processing of emotions such as fear. However, Cao argued that these null-findings may have been due to the analysis techniques that were used.

Using graph theoretical analyses, in which regions of the brain are treated as nodes in a functional network, Cao and colleagues explored the functional connectivity of unaffected siblings of schizophrenia patients and unrelated control participants while they completed a matching task involving faces showing different emotions. Interestingly, there were no differences between these groups in terms of global network organization; however, Cao was able to identify a subnetwork of limbic system and visual cortex areas in which connectivity was lower in the siblings of schizophrenia patients.

Cao was then able to to collapse this network into a single variable by averaging the strength of each connection in the network. This variable correlated with neuroticism and anxiety, in that those with lower connectivity in this subnetwork demonstrated higher levels of these schizophrenia-associated symptoms.

Similar to the findings of Collin’s group, these findings indicate a potential intermediate phenotype of limbic system connectivity, which relates to genetic risk for schizophrenia. Interestingly, Cao’s future plans include exploring the connectivity of this limbic/visual system subnetwork in patients with schizophrenia, which I actually think should have been done before looking at unaffected siblings.

[This post was originally published at my previous blog, Neurolore.]

Today’s Theme H symposium on gender bias in neuroscience was both eye-opening and frustrating at the same time. It’s clear that these biases are not only highly prevalent, but they are also directly affecting gender distribution in neuroscience careers. For example, the symposium chair, Jennifer Raymond from Stanford, noted that at the current rates of promotion in neuroscientific careers we can expect to achieve 50% of assistant professors as women by the year 2117. How depressing is that? Raymond noted that this bias is hurting innovation and excellence in neuroscience, as we are only drawing from half our potential talent pool.

The first presenter, Hannah Valantine from the Stanford School of Medicine, discussed the possibility of using institution-wide intervention to reduce the stereotyping of women in the medical fields. Her work stems from the increasing attrition of women from medical positions at higher levels of leadership. Valantine emphasizes that stereotypes often operate outside our consciousness. For instance, if elementary school students are asked to draw a scientist, greater than 50% of the time they will draw a white male. Valantine also cited the famous CV study, in which university psychology professors were asked to evaluate identical CVs, except one had the name “Karen” at the top and the other had the name “Brian.” Perhaps unsurprisingly, the professors preferred to hire Brian over Karen.

At Stanford, Valantine is developing a program  geared toward reversing this bias. The Recruitment to Expand Diversity and Excellence (REDE) program works to increase awareness of potential bias and stereotypes. Importantly, this intervention seems to decrease the strength of one’s belief in having a lack of personal bias. Additionally, a belongingness intervention at Stanford has been shown to increase belief in career advancement and personal potential among women in medicine. Overall, short term intervention seems to be capable of reducing gender bias and promoting resilience of women in the workplace.

I snuck out of the symposium to catch the talk about which I wrote my previous post so I missed the second presenter. The next presenter, Peter Glick from Lawrence University, discussed the presence of “benevolent sexism” (with the clever acronym of “BS”) in science. This type of sexism involves subtle, patronizing discrimination, as opposed to outright hostile sexism. For example, women tend to receive more positive feedback, yet are given less challenging assignments, creating a sense of hallow praise.

Benevolent sexism creates a “double bind,” in which women who experience it perform poorly (in an experimental setting), and yet if they reject this treatment, they are perceived as less warm. In other words, if you are patronized and stand up to it, you lose respect. This pattern of sexist behavior in the workplace clearly needs to change. Glick stressed the importance of recognizing this more under-the-radar type of sexism in science. To work against the effects of benevolent sexism, he proposes “wise mentoring,” in which female students are given equal amounts of critical feedback as male students. In line with Valantine’s ideas, Glick believes this feedback should be supplemented with personal assurance and encouragement, which promotes belongingness, as opposed to rejection, among female students.

The last presenter, Muriel Niederle from Stanford, has been exploring the “opt out phenomenon,” in which women are increasingly choosing not to pursue careers in the sciences. She presented data on identifying a non-cognitive skill that might explain the strong presence of gender bias in the STEM fields. Using an economics-based paradigm, Niederle has discovered that competitiveness may account for a large portion of gender bias. She has tested this idea in 9th grade classes in the Netherlands, in which she found that the degree of competitiveness among girls strongly predicts whether or not they enter academic tracks that are viewed as more prestigious, such as technology and biology. Identifying this potential moderator of gender bias will hopefully help in efforts to reverse the effects of bias on the future of neuroscience.

​[This post was originally published at my previous blog, Neurolore.]

Day 2 has already been jam packed with lots of exciting findings presented here at Neuroscience 2013. In some weird, unintentional effort to wear myself out (what is wrong with me?) I’ve been going up and down the escalators, bouncing between the poster hall and talks all morning.

In one of my forays upstairs, I sat in on a short talk by Alice Graham from the University of Oregon. Graham presented a novel way to look at a commonly studied topic: early life stress. Citing a study that discovered obesity in individuals who experienced famine in utero as opposed to postnatally, Graham emphasized that there is a mismatch between pre- and postnatal environments when it comes to risk for disease. Instead of looking at stress at a specific time point in development (e.g. only prenatal), Graham proposed looking at the change in stress that occurs as an infant transitions from the pre- to postnatal environment.

This change in stress might be particularly evident in a variable such as interparental conflict. This variable, tested via response to an angry-toned voice, has been previously shown in Graham’s work to influence functional connectivity between the medial prefrontal cortex (mPFC) and other brain regions in infants. In a new study, Graham used self-report measures to show that interparental conflict is higher in the postnatal environment, meaning after a baby is born, as compared to the prenatal environment, meaning during pregnancy. Perhaps this has something to do with sleep deprivation and the screaming, spitting up bundle of joy, I’m not sure.

Since there is a difference in the stress due to interparental conflict that an infant might experience prenatally versus postnally, there may be differential effects of this stress on development. Graham studied this difference within the context of resting-state functional connectivity in the default mode network (DMN), a functional network in the brain that is active when someone is not specifically engaged in a task. Graham chose this network because it seems to be impacted by early life stress and it exhibits rapid development from 0 to 2 years of life.

When subtracting prenatal interparental conflict from postnatal interparental conflict, Graham showed that functional connectivity was increased between the posterior cingulate cortex and the following regions: mPFC, medial temporal lobe, inferior temporal gyrus, and the right amygdala. Thus, the postnatal increase in the stress an infant might experience due to interparental conflict seems to increase functional connectivity between several regions of the DMN, effectively speeding up the development of this network.
​
Perhaps this is an adaptive response to stress. However, Graham noted that another study has reported increased DMN functional connectivity in full term infants as compared to premature infants, meaning the less stressed, full term infants actually displayed greater DMN connectivity. This finding is in contrast to the findings of Graham’s group. Accordingly, further study of the effects of change in stress across development is necessary. Perhaps different types of stress have different effects at different times. As with most questions in neuroscience, this idea is compelling, yet extremely complicated.

​[This post was originally published at my previous blog, Neurolore.]

Over in the theme H section of the gargantuan poster hall, in addition to running into one of my favorite undergraduate professors from Baylor (Sic ’em bears!), I got to check out some interesting posters.

I stopped by a poster (20.01SA) on the left-sided bias in brain depictions by Jeffery Wilson and his undergraduate students from Albion College. Think about drawing a picture of a brain on a chalkboard. Did you imagine drawing the left hemisphere of the brain? Apparently, so does everyone else. With a Google image search, Wilson and his students showed that most of the time depictions of the brain demonstrate this left-sided bias. Interestingly, this bias was not present when the search was restricted to only historical images. It’s not clear why we seem to have developed this laterality bias over time. Wilson hypothesizes that the handedness of the drawer or the left to right nature of English writing may play a part, but this random yet intriguing phenomenon deserves a closer look.

I also got to check out a poster (23.17SA) by the creators of Brain Busters, an outreach program geared toward breaking down common myths about the brain and increasing neuroscience literacy among educators and the general public. This online (free!) training program is composed of modules in which neuroscience misconceptions are confronted and discussed in an informative yet encouraging manner, leaving learners not only with a new understanding of the facts but also the ability to think critically about future neuroscience myths they may encounter. Check out this newly launching program at www.brainbusters.org. You can even buy some awesome t-shirts to support their cause.
​
Lastly, I stopped by a poster (20.16SA) by S. Robert Snodgrass from the Harbor-UCLA Medical Center. He shared some interesting ideas on the utility of animal models of human diseases and disorders. Discussing models of stroke, Alzheimer’s disease, and autism specifically, he noted that what works in an animal model, either in terms of mechanism or treatment, often doesn’t translate into human clinical studies. He cited the oversimplification of human disorders and hype over “trendy” animal models as some of the causes of this translation separation. As far as suggestions for working around this problem, Snodgrass didn’t have much to offer. Perhaps acknowledging the issue is the first step.

[This post was originally published at my previous blog, Neurolore.]

It’s only been Day 1 of Neuroscience 2013, I am already filling up with awesome findings to share (plus my feet are already mad at me). Since my themes are pretty disparate (C: Disorders of the Nervous System & H: History/Teaching/Public Awareness/Societal Impacts) I figured it would be helpful to post on each theme separately throughout the week. So here we go:

SfN Day 1/Theme C – Human Biomarkers of Autism
​
The first poster I visited turned out to be very relevant to my work as well as pretty interesting. Annette Ye of the University of Toronto took the time to discuss her poster (49.01) on the highly explored topic of altered functional connectivity in adolescents with autism spectrum disorder (ASD), only she looked at this idea using magnetoencephalography (MEG), as opposed to functional magnetic resonance imaging (fMRI). MEG data complements the information provided by fMRI, as it provides a picture of the synchrony of oscillations in neural activity. Using resting state MEG data, Ye and colleagues explored the synchrony between 90 brain regions using different frequency bands. When comparing individuals with ASD to those without ASD, they found several regions, most of which were in the frontal lobe, that exhibited increased inter-regional connectivity in the gamma frequency band. This frequency band is related to local as opposed to long range connections and may be predictive of cognition. Disruption in neural synchrony within this band has been associated with schizophrenia and depression.

When Ye used graph theoretical analysis, in which brain regions are treated as nodes in a connected functional network, the increased functional connectivity among the frontal lobe regions was confirmed by increased node strength and clustering coefficients, indicating stronger inter-regional connections and a greater likelihood of neighboring connections, respectively. These MEG findings replicate previous findings that functional connectivity may be higher in ASD in local areas of the brain, including those related to cognition. Whereas decreased long-range connectivity in ASD has been robustly demonstrated in current literature, the local hyper-connectivity argument has been less defined. However, the findings Ye presented show that this argument definitely warrants further consideration.

Ye’s next steps for this project include collecting more data so she can explore potential relationships between this increased frontal lobe connectivity and the severity of autism symptoms, such as impairments in social communication or restricted, repetitive behaviors. I would guess that this increased frontal lobe connectivity would be related to increased repetitive behaviors, such as resistance to change. I look forward to what Ye finds.

[This post was originally published at my previous blog, Neurolore.]