The default mode network (DMN) changes with Alzheimer’s disease (AD), mild cognitive impairment (MCI), and normal aging, but these changes are not well characterized. The posterior cingulate cortex (PCC) is a critical hub in the DMN. Using data from the Alzheimer’s Disease Neuroimaging Initiative database, we modeled longitudinal change in PCC connectivity in three subject groups: healthy controls, MCI, and AD, and tested whether connectivity changes were linear or quadratic with time. The HC group had no evidence of Aβ accumulation (N=56, 25 male, 75.7±6.23 years). The MCI group pooled all MCI and significant memory concern subjects with no signs of dementia (N=127, 59 male, 72.6±7.14 years). The AD group fulfilled the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS) and the Alzheimer's Disease and Related Disorders Association (ADRDA, now known as the Alzheimer’s Association) criteria for probable AD (N=34, 16 male, 74.2±7.47 years). We processed resting state fMRI images using a workflow that represents best practices for removal of noise, censoring motion outliers. We measured individual connectivity to a 10mm sphere in the PCC (2,-52,28) and transformed the Pearson correlations using a Fisher Z-transformation. We used a software developed in our lab called neuropointillist to fit a mixed effects growth model in R to each voxel, comparing linear and quadratic forms of longitudinal change. Data were cluster corrected for multiple comparisons at p=.05, α=.05. A linear model was superior to a quadratic model for all groups across the brain. All three groups showed decreased connectivity with age of the PCC to the medial frontal wall. However, we found increased connectivity to the dorsal attention network only in the MCI group. We identified distinct regions of the brain that increased and decreased in connectivity to the PCC with age in the three subgroups. These differences likely represent the different pathophysiological processes occurring in normal aging, MCI, and AD.

Insect flight relies heavily on visual sensing. In many flight behaviors (e.g. navigating over long distances or through cluttered environments, finding food sources, or evading predators), insects must parse the visual scene to extract an estimate of their own motion and identify external objects or agents moving in their environment. Across numerous taxa and behaviors, there is a rich literature exploring behavioral responses to wide-field optic flow (visual stimuli arising from egomotion) and small-field target motion (cues corresponding to exogenous motion), primarily in the yaw dynamics involved in navigation. In contrast to yaw which is marginally stable, the equilibrium about pitch angle is inherently unstable, hence there are significant consequences to adjusting the flight attitude. To stabilize the visual scene under this constraint, insects can either reorient their body or move their head to redirect gaze. In this work, we investigate how the hawkmoth, Manduca sexta, modulates body pitch and gaze angle in response to wide- and small-field visual motion. Moths are tethered to a freely rotating armature at the center of a cylindrical arena and presented an image of a circular flower against a background grating. Figure and ground are oscillated both individually as well as simultaneously (both synchronously and incongruously). A multi-input--multi-output analysis reveals correlations that suggest moths employ parallel strategies for stabilizing posture and gaze dependent on the spatio-temporal content of the visual scene. The inherent instability in pitch dynamics necessitates these dual strategies. Distinguishing between and reacting to wide- and small-field visual stimuli are necessary parts of many animal behaviors; through this research we hope to better understand how biological systems assimilate motion and mechanical cues to coordinate effective movements.

In mammals, daily rhythms of physiology and behavior are synchronized to a 24-hour light-dark (LD) cycle via input to the suprachiasmatic nucleus (SCN), the master circadian clock. Disruptions to the circadian system are associated with neurological and psychiatric disorders, including mood disorders like depression. These disorders are associated with symptoms such as sexual dysfunction and anhedonia, the inability to experience pleasure from activities that are typically enjoyable. Previous work has shown that rats exposed to a 22-hour LD cycle, which leads to desynchrony between neuronal oscillators in the SCN and mimics an internally desynchronized state, display behaviors indicative of these symptoms. Jet lag is a common challenge to the circadian system and, in rats, causes desynchrony between SCN neuronal oscillators. We hypothesized that desynchrony induced by jet lag could also lead to depressive symptoms. We measured sexual behavior before and after a 6-hour phase shift, simulating jet lag. Male rats were placed in an arena with a receptive female and sexual activity was quantified. We also performed a saccharin preference test where males had continuous access to water and saccharin solution. The volume drank from each bottle was measured daily; anhedonia is evident as a decrease in the preference for saccharin solution. Locomotor activity was recorded using infrared detectors, to associate behavior with the amount of time it took for males to re-entrain to the new LD cycle. Six-hour advances and delays of the LD cycle were not associated with behavioral manifestations of depression. This work, combined with previous work, provides greater insight into the consequences of circadian disruption and suggests that mental health consequences depend on the specific type of circadian misalignment. Whereas steady state misalignment induced by exposure to 22-hour LD cycles is associated with a depressive phenotype, the transient misalignment induced by jet lag is not.

The research project primarily aims at better characterizing how neurons of certain types transform their inputs from “upstream” neurons into outputs to “downstream” neurons. So far in the project, it has been determined that there are major differences between Intratelencephalic (IT) and Pyramidal Tract (PT) neurons on the basis of genetics, morphology, electrophysiology and functionality. We have collected data from two genetically distinct mice. Single cell recordings reveal that PT and IT type neurons are electro physiologically different, in both the locations to which they send their signals in the nervous system and the neuron morphology. Using retrograde tracing (a research method used to trace neural connections), it was found out that PT type send their axons to spinal cord and regions outside of telencephalon region of the brain but IT were restricted to telencephalon region of the brain. Now, my primary focus will be to elucidate how the electrophysiological and morphological characteristics interact. To do this, I will use reconstructions from two photon imaged stacks of their structure imported into a simulation environment called NEURON. Using NEURON, I will try to identify the potential mechanisms that underlines the sensitivity in synaptic integration of the two neuron types. This will provide testable hypothesis for future experimentation. IT and PT type have distinct roles. IT type are sensitive to anti-depressing mechanisms whereas PT type are sensitive to ALS mechanisms. It is this property that will help us create targeted medications.

The human brain is comprised of many billions of neurons, which then connect with each other many trillions of times. Modeling how these neurons function provides insight into language, thought, and behavior. However, neurons are not all identical, so an approach that performs well on some neurons will perform poorly on others. I focused on a learning system to detect and model neurons that use pattern-recognition in order to decide when to fire, which would not be expressed well under existing approaches. We researched a strategy for describing these process in a non-Markov state, which means that we do not have enough information to correctly model the neuron, by approximating to Markov models using a subset of the required information. This means that we produce a set of probabilities for the next state of the neuron given the previous state. We approximate the Markov model using Monte-Carlo Tree Search (MCTS) optimizing for smallest confidence interval to select sequences to measure. In order to calculate a confidence interval on a given sequence we apply A*, pronounced A star, which is a targeted pathfinding algorithm, in order to produce execution paths that in turn find confidence intervals. We use two strategies for pruning sequences; a MCTS to find the measured sequence that differs least from chance for long term learning, and exponential decay for short term learning. We assume that after we process information it cannot be recovered, so when we measure new sequences we only measure from the time of consideration onward. If our process successfully locates important patterns the neuron is looking for, it produces an accurate approximation for the neuron that can be rapidly evaluated.