The functional characterization of neuronal subtypes within in vitro models is critical for understanding neural circuit development. While traditional wet lab methods provide structural ground truth, high throughput microelectrode array (MEA) electrophysiology offers dynamic insights into network activity. The objective of this study was to use spike waveform features to classify putative excitatory versus inhibitory associated electrophysiological signatures and to evaluate how a multimodal AI model (Gemini) can assist with waveform based cell characterization using a WaveMAP style dimensionality reduction and graph clustering pipeline. Using MEA data from cortex (Cx) organoids and microglia+cortex (M+C) assembloids, the AI assisted pipeline utilized t-SNE projection and agglomerative clustering to isolate distinct morphological signatures. The biological data revealed an inhibitory dominant recording in Cx models, compared to M+C model that showed a physiologically close excitatory to inhibitory (E:I) ratio. However, critical algorithmic limitations in Gemini’s clustering highlights the need for spatial transcriptomics or other experimental methods to corroborate neuron identity. While AI driven waveform clustering is highly interactive, it is insufficient as a standalone tool for definitive network labeling without molecular validation.

Glioblastoma (GBM) is a fatal form of brain cancer. Standard care treatments such as resection, radiation, and chemotherapy have only marginally extended life expectancy. Our lab seeks to characterize the GBM cells left after resection in the non-enhancing (NE) region and their interactions with non-transformed cells. In our single-nucleus RNA-sequencing (snRNA-seq) dataset of biopsies containing both the traditionally resected contrast-enhancing (CE) region and NE regions, we found an enrichment of oligodendrocyte progenitor cells (OPCs) in the NE region compared to the CE region. Our predicted cell-cell interactions from this dataset led us to hypothesize that OPCs and GBM cells influence each other competitively. Thus, we tested the effects of immature OPC-conditioned media (CM) on GBM cell count and vice versa. After one week in culture, the OPC CM condition resulted in decreased GBM cell number, while the GBM CM condition resulted in increased OPC cell number. These data support the hypothesis that immature OPCs express factors to enhance their potential competitive fitness, suggesting a GBM-responsive increase in OPC proliferation within the tumor microenvironment. With this study, we hope to gain a deeper understanding of cellular interactions within the NE tumor microenvironment, which may lead to the identification of future therapeutic targets and the development of novel treatments that manipulate this competitive process to inhibit invasion in the tumor microenvironment during later stages of GBM.

Understanding the impact of social connectedness on brain aging is essential to combat social factors of brain aging. This longitudinal study examined the relationship between social connectedness and volume of the right para-insular region (R_PI_ROI) in 914 healthy participants (ages 36-90+, mean 61.29) from the Aging Adult Brain Connectome (AABC) study. Participants were assessed in datapoints two years apart. Three indicators (loneliness, social satisfaction, emotional support) were analyzed using the NIH Toolbox Assessment. Changes in the right parainsular volume and social scores were compared via robust linear regression in R. The analysis controlled for age, sex, and education, and p-values were corrected via FDR. Results showed significant negative relationships for social satisfaction (p = 6.33×10⁻⁷) and emotional support (p = 6.17×10⁻⁵) with changes in R_PI_ROI, indicating that declines in support/satisfaction are associated with greater volume decline. Increased loneliness was also significantly associated with R_PI_ROI decline (p = 8.78E-06). Specifically, a one-point increase in loneliness score corresponded to a 7.064571 decrease in R_PI_ROI volume. Social connectedness is associated with decreasing para-insular region volume, suggesting these lifestyle factors require increased emphasis for aging adults. These results demonstrate the importance for public health campaigns addressing brain aging to focus on social factors.

The UCLA BRAIN Trial investigates how personalized transcranial magnetic stimulation (TMS) targeting an overactive prefrontal–amygdala circuit promotes recovery for chronic post-concussive symptoms following mild-traumatic brain injury (mTBI). However, the contribution of social determinants of health (SDOH) to fear avoidance behaviors and post-concussive outcomes remains unclear. This study characterizes socioeconomic and psychosocial disadvantage using the Area Deprivation Index (ADI), Social Vulnerability Index (SVI), and perceived social support (PROMIS), examining their relationships with symptom severity, fear-avoidance behavior, and degree of symptom change following treatment. SDOH distributions were characterized using descriptive statistics and normality testing. Then, correlations between SDOH variables and clinical outcomes using Spearman's coefficients were examined. Perceived social support showed a trending positive association with symptom improvement (r = 0.27, p = .07), suggesting that individuals with stronger social networks experienced more favorable recovery trajectories. Neighborhood-level factors (SVI, ADI) showed weak associations with baseline symptoms (r ≈ 0.08), likely reflecting low socioeconomic variance in this sample. These findings suggest that psychosocial factors, particularly social support, may be more influential than neighborhood disadvantage in predicting mTBI recovery, warranting integration of social screening into concussion care.

Neural polarization, a phenomenon in which individuals exhibit different patterns of brain activity in response to the same stimulus, has been observed in political contexts, where it is associated with differences in prior beliefs (e.g. conservative vs liberal). However, it remains unclear whether similar effects arise in other domains shaped by strong individual preferences, such as music. Unlike politics, music is widely recognized for its ability to foster emotional connection and may offer a unique pathway to reduce, rather than reinforce, neural divergence. This study investigates whether shared musical preferences are associated with increased neural synchrony and whether guided exposure to unfamiliar genres can reduce neural polarization. Fifty adults (25 country and 25 pop music enthusiasts) will undergo fMRI while listening to excerpts from preferred, non-preferred, and neutral genres. Neural synchrony will be quantified using inter-subject correlation analysis, and post-scan surveys will assess familiarity and emotional engagement. Participants showing the greatest neural divergence will then complete guided exposure to their non-preferred genre, followed by a second fMRI session. We hypothesize that shared musical preferences will correspond to greater neural synchrony, and that guided exposure will increase cross-group synchrony and enjoyment of unfamiliar music. These findings aim to establish a biological basis for music’s role in aligning neural responses, with implications for reducing bias.