Author name cluster

Yi Qi

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

2 author rows

YNIMG Journal 2024 Journal Article

High-resolution diffusion magnetic resonance imaging and spatial-transcriptomic in developing mouse brain

Xinyue Han
Surendra Maharjan
Jie Chen
Yi Zhao
Yi Qi
Leonard E. White
G. Allan Johnson
Nian Wang

Brain development is a highly complex process regulated by numerous genes at the molecular and cellular levels. Brain tissue exhibits serial microstructural changes during the development process. High-resolution diffusion magnetic resonance imaging (dMRI) affords a unique opportunity to probe these changes in the developing brain non-destructively. In this study, we acquired multi-shell dMRI datasets at 32 µm isotropic resolution to investigate the tissue microstructure alterations, which we believe to be the highest spatial resolution dMRI datasets obtained for postnatal mouse brains. We adapted the Allen Developing Mouse Brain Atlas (ADMBA) to integrate quantitative MRI metrics and spatial transcriptomics. Diffusion tensor imaging (DTI), diffusion kurtosis imaging (DKI), and neurite orientation dispersion and density imaging (NODDI) metrics were used to quantify brain development at different postnatal days. We demonstrated that the differential evolutions of fiber orientation distributions contribute to the distinct development patterns in white matter (WM) and gray matter (GM). Furthermore, the genes enriched in the nervous system that regulate brain structure and function were expressed in spatial correlation with age-matched dMRI. This study is the first one providing high-resolution dMRI, including DTI, DKI, and NODDI models, to trace mouse brain microstructural changes in WM and GM during postnatal development. This study also highlighted the genotype-phenotype correlation of spatial transcriptomics and dMRI, which may improve our understanding of brain microstructure changes at the molecular level.

Details DOI

ICML Conference 2024 Conference Paper

Position: Building Guardrails for Large Language Models Requires Systematic Design

Yi Dong 0002
Ronghui Mu
Gaojie Jin
Yi Qi
Jinwei Hu
Xingyu Zhao 0001
Jie Meng
Wenjie Ruan

As Large Language Models (LLMs) become more integrated into our daily lives, it is crucial to identify and mitigate their risks, especially when the risks can have profound impacts on human users and societies. Guardrails, which filter the inputs or outputs of LLMs, have emerged as a core safeguarding technology. This position paper takes a deep look at current open-source solutions (Llama Guard, Nvidia NeMo, Guardrails AI), and discusses the challenges and the road towards building more complete solutions. Drawing on robust evidence from previous research, we advocate for a systematic approach to construct guardrails for LLMs, based on comprehensive consideration of diverse contexts across various LLMs applications. We propose employing socio-technical methods through collaboration with a multi-disciplinary team to pinpoint precise technical requirements, exploring advanced neural-symbolic implementations to embrace the complexity of the requirements, and developing verification and testing to ensure the utmost quality of the final product.

Details

YNIMG Journal 2022 Journal Article

Resolution and b value dependent structural connectome in ex vivo mouse brain

Stephanie Crater
Surendra Maharjan
Yi Qi
Qi Zhao
Gary Cofer
James C. Cook
G. Allan Johnson
Nian Wang

Diffusion magnetic resonance imaging has been widely used in both clinical and preclinical studies to characterize tissue microstructure and structural connectivity. The diffusion MRI protocol for the Human Connectome Project (HCP) has been developed and optimized to obtain high-quality, high-resolution diffusion MRI (dMRI) datasets. However, such efforts have not been fully explored in preclinical studies, especially for rodents. In this study, high quality dMRI datasets of mouse brains were acquired at 9. 4T system from two vendors. In particular, we acquired a high-spatial resolution dMRI dataset (25 μm isotropic with 126 diffusion encoding directions), which we believe to be the highest spatial resolution yet obtained; and a high-angular resolution dMRI dataset (50 μm isotropic with 384 diffusion encoding directions), which we believe to be the highest angular resolution compared to the dMRI datasets at the microscopic resolution. We systematically investigated the effects of three important parameters that affect the final outcome of the connectome: b value (1000s/mm2 to 8000 s/mm2), angular resolution (10 to 126), and spatial resolution (25 µm to 200 µm). The stability of tractography and connectome increase with the angular resolution, where more than 50 angles is necessary to achieve consistent results. The connectome and quantitative parameters derived from graph theory exhibit a linear relationship to the b value (R2 > 0. 99); a single-shell acquisition with b value of 3000 s/mm2 shows comparable results to the multi-shell high angular resolution dataset. The dice coefficient decreases and both false positive rate and false negative rate gradually increase with coarser spatial resolution. Our study provides guidelines and foundations for exploration of tradeoffs among acquisition parameters for the structural connectome in ex vivo mouse brain.

Details DOI

YNIMG Journal 2020 Journal Article

Cytoarchitecture of the mouse brain by high resolution diffusion magnetic resonance imaging

Nian Wang
Leonard E. White
Yi Qi
Gary Cofer
G. Allan Johnson

MRI has been widely used to probe the neuroanatomy of the mouse brain, directly correlating MRI findings to histology is still challenging due to the limited spatial resolution and various image contrasts derived from water relaxation or diffusion properties. Magnetic resonance histology has the potential to become an indispensable research tool to mitigate such challenges. In the present study, we acquired high spatial resolution MRI datasets, including diffusion MRI (dMRI) at 25 μm isotropic resolution and quantitative susceptibility mapping (QSM) at 21.5 μm isotropic resolution to validate with conventional mouse brain histology. Diffusion weighted images (DWIs) show better delineation of cortical layers and glomeruli in the olfactory bulb than fractional anisotropy (FA) maps. However, among all the image contrasts, including quantitative susceptibility mapping (QSM), T1/T2∗ images and DTI metrics, FA maps highlight unique laminar architecture in sub-regions of the hippocampus, including the strata of the dentate gyrus and CA fields of the hippocampus. The mean diffusivity (MD) and axial diffusivity (AD) yield higher correlation with DAPI (0.62 and 0.71) and NeuN (0.78 and 0.74) than with NF-160 (-0.34 and -0.49). The correlations between FA and DAPI, NeuN, and NF-160 are 0.31, -0.01, and -0.49, respectively. Our findings demonstrate that MRI at microscopic resolution deliver a three-dimensional, non-invasive and non-destructive platform for characterization of fine structural detail in both gray matter and white matter of the mouse brain.

Details DOI

YNIMG Journal 2020 Journal Article

Variability and heritability of mouse brain structure: Microscopic MRI atlases and connectomes for diverse strains

Nian Wang
Robert J. Anderson
David G. Ashbrook
Vivek Gopalakrishnan
Youngser Park
Carey E. Priebe
Yi Qi
Rick Laoprasert

Genome-wide association studies have demonstrated significant links between human brain structure and common DNA variants. Similar studies with rodents have been challenging because of smaller brain volumes. Using high field MRI (9. 4 T) and compressed sensing, we have achieved microscopic resolution and sufficiently high throughput for rodent population studies. We generated whole brain structural MRI and diffusion connectomes for four diverse isogenic lines of mice (C57BL/6J, DBA/2J, CAST/EiJ, and BTBR) at spatial resolution 20, 000 times higher than human connectomes. We measured narrow sense heritability (h2 ) I. e. the fraction of variance explained by strains in a simple ANOVA model for volumes and scalar diffusion metrics, and estimates of residual technical error for 166 regions in each hemisphere and connectivity between the regions. Volumes of discrete brain regions had the highest mean heritability (0. 71 ± 0. 23 SD, n = 332), followed by fractional anisotropy (0. 54 ± 0. 26), radial diffusivity (0. 34 ± 0. 022), and axial diffusivity (0. 28 ± 0. 19). Connection profiles were statistically different in 280 of 322 nodes across all four strains. Nearly 150 of the connection profiles were statistically different between the C57BL/6J, DBA/2J, and CAST/EiJ lines. Microscopic whole brain MRI/DTI has allowed us to identify significant heritable phenotypes in brain volume, scalar DTI metrics, and quantitative connectomes.

Details DOI

NeurIPS Conference 2018 Conference Paper

Bandit Learning with Implicit Feedback

Yi Qi
Qingyun Wu
Hongning Wang
Jie Tang
Maosong Sun

Implicit feedback, such as user clicks, although abundant in online information service systems, does not provide substantial evidence on users' evaluation of system's output. Without proper modeling, such incomplete supervision inevitably misleads model estimation, especially in a bandit learning setting where the feedback is acquired on the fly. In this work, we perform contextual bandit learning with implicit feedback by modeling the feedback as a composition of user result examination and relevance judgment. Since users' examination behavior is unobserved, we introduce latent variables to model it. We perform Thompson sampling on top of variational Bayesian inference for arm selection and model update. Our upper regret bound analysis of the proposed algorithm proves its feasibility of learning from implicit feedback in a bandit setting; and extensive empirical evaluations on click logs collected from a major MOOC platform further demonstrate its learning effectiveness in practice.

PDF Details

YNIMG Journal 2016 Journal Article

The fornix provides multiple biomarkers to characterize circuit disruption in a mouse model of Alzheimer's disease

Alexandra Badea
Lauren Kane
Robert J. Anderson
Yi Qi
Mark Foster
Gary P. Cofer
Neil Medvitz
Anne F. Buckley

Multivariate biomarkers are needed for detecting Alzheimer's disease (AD), understanding its etiology, and quantifying the effect of therapies. Mouse models provide opportunities to study characteristics of AD in well-controlled environments that can help facilitate development of early interventions. The CVN-AD mouse model replicates multiple AD hallmark pathologies, and we identified multivariate biomarkers characterizing a brain circuit disruption predictive of cognitive decline. In vivo and ex vivo magnetic resonance imaging (MRI) revealed that CVN-AD mice replicate the hippocampal atrophy (6%), characteristic of humans with AD, and also present changes in subcortical areas. The largest effect was in the fornix (23% smaller), which connects the septum, hippocampus, and hypothalamus. In characterizing the fornix with diffusion tensor imaging, fractional anisotropy was most sensitive (20% reduction), followed by radial (15%) and axial diffusivity (2%), in detecting pathological changes. These findings were strengthened by optical microscopy and ultrastructural analyses. Ultrastructual analysis provided estimates of axonal density, diameters, and myelination—through the g-ratio, defined as the ratio between the axonal diameter, and the diameter of the axon plus the myelin sheath. The fornix had reduced axonal density (47% fewer), axonal degeneration (13% larger axons), and abnormal myelination (1. 5% smaller g-ratios). CD68 staining showed that white matter pathology could be secondary to neuronal degeneration, or due to direct microglial attack. In conclusion, these findings strengthen the hypothesis that the fornix plays a role in AD, and can be used as a disease biomarker and as a target for therapy.

Details DOI

YNIMG Journal 2010 Journal Article

Ultrasonic disruption of the blood–brain barrier enables in vivo functional mapping of the mouse barrel field cortex with manganese-enhanced MRI

Gabriel P. Howles
Yi Qi
G. Allan Johnson

Though mice are the dominant model system for studying the genetic and molecular underpinnings of neuroscience, functional neuroimaging in mice remains technically challenging. One approach, Activation-Induced Manganese-enhanced MRI (AIM MRI), has been used successfully to map neuronal activity in rodents. In AIM MRI, manganese2+ acts a calcium analog and accumulates in depolarized neurons. Because manganese2+ shortens T1, regions of elevated neuronal activity enhance in MRI. However, because manganese does not cross the blood–brain barrier (BBB), the need to osmotically disrupt the BBB has limited the use of AIM MRI, particularly in mice. In this work, the BBB was opened in mice using unfocused, transcranial ultrasound in combination with gas-filled microbubbles. Using this noninvasive technique to open the BBB bilaterally, manganese could be quickly administered to the whole mouse brain. With this approach, AIM MRI was used to map the neuronal response to unilateral mechanical stimulation of the vibrissae in lightly sedated mice. The resultant 3D activation map agreed well with published representations of the vibrissae regions of the barrel field cortex. The anterior portions of the barrel field cortex corresponding to the more rostral vibrissae showed greater activation, consistent with previous literature. Because the ultrasonic opening of the BBB is simple, fast, and noninvasive, this approach is suitable for high-throughput and longitudinal studies in awake mice. This approach enables a new way to map neuronal activity in mice with manganese.

Details DOI