Background: Health care providers and health-related researchers face significant challenges when applying sentiment analysis tools to health-related free-text survey data. Most state-of-the-art applications were developed in domains such as social media, and their performance in the health care context remains relatively unknown. Moreover, existing studies indicate that these tools often lack accuracy and produce inconsistent results. Objective: This study aims to address the lack of comparative analysis on sentiment analysis tools applied to health-related free-text survey data in the context of COVID-19. The objective was to automatically predict sentence sentiment for 2 independent COVID-19 survey data sets from the National Institutes of Health and Stanford University. Methods: Gold standard labels were created for a subset of each data set using a panel of human raters. We compared 8 state-of-the-art sentiment analysis tools on both data sets to evaluate variability and disagreement across tools. In addition, few-shot learning was explored by fine-tuning Open Pre-Trained Transformers (OPT; a large language model [LLM] with publicly available weights) using a small annotated subset and zero-shot learning using ChatGPT (an LLM without available weights). Results: The comparison of sentiment analysis tools revealed high variability and disagreement across the evaluated tools when applied to health-related survey data. OPT and ChatGPT demonstrated superior performance, outperforming all other sentiment analysis tools. Moreover, ChatGPT outperformed OPT, exhibited higher accuracy by 6% and higher F-measure by 4% to 7%. Conclusions: This study demonstrates the effectiveness of LLMs, particularly the few-shot learning and zero-shot learning approaches, in the sentiment analysis of health-related survey data. These results have implications for saving human labor and improving efficiency in sentiment analysis tasks, contributing to advancements in the field of automated sentiment analysis.
Extensive hierarchical yet highly reciprocal interactions among cortical areas are fundamental for information processing. However, connectivity rules governing the specificity of such corticocortical connections, and top-down feedback projections in particular, are poorly understood. We analyze synaptic strength from functionally relevant brain areas to diverse neuronal types in the primary somatosensory cortex (S1). Long-range projections from different areas preferentially engage specific sets of GABAergic neurons in S1. Projections from other somatosensory cortices strongly recruit parvalbumin (PV)-positive GABAergic neurons and lead to PV neuron-mediated feedforward inhibition of pyramidal neurons in S1. In contrast, inputs from whisker-related primary motor cortex are biased to vasoactive intestinal peptide (VIP)-positive GABAergic neurons and potentially result in VIP neuron-mediated disinhibition. Regardless of the input areas, somatostatin-positive neurons receive relatively weak long-range inputs. Computational analyses suggest that a characteristic combination of synaptic inputs to different GABAergic IN types in S1 represents a specific long-range input area.
In order to obtain the sample sizes needed for robustly reproducible effects, it is often necessary to acquire data at multiple sites using different MRI scanners. This poses a challenge for investigators to account for the variance due to scanner, as balanced sampling is often not an option. Similarly, longitudinal studies must deal with known and unknown changes to scanner hardware and software over time. In this manuscript, we have explored scanner-related differences in the dataset recently released by the Adolescent Brain Cognitive Development (ABCD) project, a multi-site, longitudinal study of children age 9-10. We demonstrate that scanner manufacturer, model, as well as the individual scanner itself, are detectable in the resting and task-based fMRI results of the ABCD dataset. We further demonstrate that these differences can be harmonized using an empirical Bayes approach known as ComBat. We argue that accounting for scanner variance, including even minor differences in scanner hardware or software, is crucial for any analysis.
The ability to perceive spatial environments and locate oneself during navigation is crucial for the survival of animals. Mounting evidence suggests a role of the medial prefrontal cortex (mPFC) in spatially related behaviors. However, the properties of mPFC spatial encoding and how it is influenced by animal behavior are poorly defined. Here, we train the mice to perform 3 tasks differing in working memory and reward-seeking: a delayed non-match to place (DNMTP) task, a passive alternation (PA) task, and a free-running task. Single-unit recording in the mPFC shows that although individual mPFC neurons exhibit spatially selective firing, they do not reliably represent the animal location. The population activity of mPFC neurons predicts the animal location. Notably, the population coding of animal locations by the mPFC is modulated by animal behavior in that the coding accuracy is higher in tasks involved in working memory and reward-seeking. This study reveals an approach whereby the mPFC encodes spatial positions and the behavioral variables affecting it.
The COVID-19 pandemic’s impact on mental health is challenging to quantify because pre-existing risk, disease burden and public policy varied across individuals, time and regions. Longitudinal, within-person analyses can determine whether pandemic-related changes in social isolation impacted mental health. We analyzed time-varying associations between psychiatric vulnerability, loneliness, psychological distress and social distancing in a US-based study during the first year of the pandemic. We surveyed 3,655 participants about psychological health and COVID-19-related circumstances every 2 weeks for 6 months. We combined self-reports with regional social distancing estimates and a classifier that predicted probability of psychiatric diagnosis at enrollment. Loneliness and psychiatric vulnerability both impacted psychological distress. Loneliness and distress were also linked to social isolation and stress associated with distancing, and psychiatric vulnerability shaped how regional distancing affected loneliness across time. Public health policies should address loneliness when encouraging social distancing, particularly in those at risk for psychiatric conditions.