Yuda Bi

Yuda Bi

Auto-generated profile, based on publicly available information

Yuda Bi

Georgia Institute of Technology, and Emory University

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 November 2024 in Human Brain Mapping
Verified
Journal Article
Open access
YBYuda Bi
Yuda Bi
AAAnees Abrol
Anees Abrol
ZFZening Fu
Zening Fu
VCVince D. Calhoun
Vince D. Calhoun
Multimodal neuroimaging is an emerging field that leverages multiple sources of information to diagnose specific brain disorders, especially when deep learning‐based AI algorithms are applied. The successful combination of different brain imaging modalities using deep learning remains a challenging yet crucial research topic. The integration of structural and functional modalities is particularly important for the diagnosis of various brain disorders, where structural information plays a crucial role in diseases such as Alzheimer's, while functional imaging is more critical for disorders such as schizophrenia. However, the combination of functional and structural imaging modalities can provide a more comprehensive diagnosis. In this work, we present MultiViT, a novel diagnostic deep learning model that utilizes vision transformers and cross‐attention mechanisms to effectively fuse information from 3D gray matter maps derived from structural MRI with functional network connectivity matrices obtained from functional MRI using the ICA algorithm. MultiViT achieves an AUC of 0.833, outperforming both our unimodal and multimodal baselines, enabling more accurate classification and diagnosis of schizophrenia. In addition, using vision transformer's unique attentional maps in combination with cross‐attentional mechanisms and brain function information, we identify critical brain regions in 3D gray matter space associated with the characteristics of schizophrenia. Our research not only significantly improves the accuracy of AI‐based automated imaging diagnostics for schizophrenia, but also pioneers a rational and advanced data fusion approach by replacing complex, high‐dimensional fMRI information with functional network connectivity, integrating it with representative structural data from 3D gray matter images, and further providing interpretative biomarker localization in a 3D structural space.

7 June 2024 in NeuroImage
Verified
Journal Article
Open access
YBYuda Bi
Yuda Bi
AAAnees Abrol
Anees Abrol
SJSihan Jia
Sihan Jia
JSJing Sui
Jing Sui
VCVince Calhoun
Vince Calhoun
Brain disorders are often associated with changes in brain structure and function, where functional changes may be due to underlying structural variations. Gray matter (GM) volume segmentation from 3D structural MRI offers vital structural information for brain disorders like schizophrenia, as it encompasses essential brain tissues such as neuronal cell bodies, dendrites, and synapses, which are crucial for neural signal processing and transmission; changes in GM volume can thus indicate alterations in these tissues, reflecting underlying pathological conditions. In addition, the use of the ICA algorithm to transform high-dimensional fMRI data into functional network connectivity (FNC) matrices serves as an effective carrier of functional information. In our study, we introduce a new generative deep learning architecture, the conditional efficient vision transformer generative adversarial network (cEViT-GAN), which adeptly generates FNC matrices conditioned on GM to facilitate the exploration of potential connections between brain structure and function. We developed a new, lightweight self-attention mechanism for our ViT-based generator, enhancing the generation of refined attention maps critical for identifying structural biomarkers based on GM. Our approach not only generates high quality FNC matrices with a Pearson correlation of 0.74 compared to real FNC data, but also uses attention map technology to identify potential biomarkers in GM structure that could lead to functional abnormalities in schizophrenia patients. Visualization experiments within our study have highlighted these structural biomarkers, including the medial prefrontal cortex (mPFC), dorsolateral prefrontal cortex (DL-PFC), and cerebellum. In addition, through cross-domain analysis comparing generated and real FNC matrices, we have identified functional connections with the highest correlations to structural information, further validating the structure-function connections. This comprehensive analysis helps to understand the intricate relationship between brain structure and its functional manifestations, providing a more refined insight into the neurobiological research of schizophrenia.

18 July 2023 in bioRxiv
Verified
Preprint
Open access
PDF file
YBYuda Bi
Yuda Bi
AAAnees Abrol
Anees Abrol
ZFZening Fu
Zening Fu
VCVince Calhoun
Vince Calhoun
Deep learning models, despite their potential for increasing our understanding of intricate neuroimaging data, can be hampered by challenges related to interpretability. Multimodal neuroimaging appears to be a promising approach that allows us to extract supplementary information from various imaging modalities. It’s noteworthy that functional brain changes are often more pronounced in schizophrenia, albeit potentially less reproducible, while structural MRI effects are more replicable but usually manifest smaller effects. Instead of conducting isolated analyses for each modality, the joint analysis of these data can bolster the effects and further refine our neurobiological understanding of schizophrenia. This paper introduces a novel deep learning model, the multimodal vision transformer (MultiViT), specifically engineered to enhance the accuracy of classifying schizophrenia by using structural MRI (sMRI) and functional MRI (fMRI) data independently and simultaneously leveraging the combined information from both modalities. This study uses functional network connectivity data derived from a fully automated independent component analysis method as the fMRI features and segmented gray matter volume (GMV) as the sMRI features. These offer sensitive, high-dimensional features for learning from structural and functional MRI data. The resulting MultiViT model is lightweight and robust, outperforming unimodal analyses. Our approach has been applied to data collected from control subjects and patients with schizophrenia, with the MultiViT model achieving an AUC of 0.833, which is significantly higher than the average 0.766 AUC for unimodal baselines and 0.78 AUC for multimodal baselines. Advanced algorithmic approaches for predicting and characterizing these disorders have consistently evolved, though subject and diagnostic heterogeneity pose significant challenges. Given that each modality provides only a partial representation of the brain, we can gather more comprehensive information by harnessing both modalities than by relying on either one independently. Furthermore, we conducted a saliency analysis to gain insights into the co-alterations in structural gray matter and functional network connectivity disrupted in schizophrenia. While it’s clear that the MultiViT model demonstrates differences compared to previous multimodal methods, the specifics of how it compares to methods such as MCCA and JICA are still under investigation, and more research is needed in this area. The findings underscore the potential of interpretable multimodal data fusion models like the MultiViT, highlighting their robustness and potential in the classification and understanding of schizophrenia.

14 November 2025
Verified
Book Chapter
Closed access
PDF file
XHXu Hu
Xu Hu
JYJingling Yang
Jingling Yang
WXWenjun Xiao
Wenjun Xiao
SJSihan Jia
Sihan Jia
YGYutong Gao
Yutong Gao
YBYuda Bi
Yuda Bi

29 July 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
JSJingnan Sun
Jingnan Sun
YBYuda Bi
Yuda Bi
ASAnruo Shen
Anruo Shen
YJYanyan Jiang
Yanyan Jiang
YSYike Sun
Yike Sun
Background: Post-COVID brain fog is clinically important but overlaps with respiratory, systemic, fatigue, sleep, and mood symptoms. Apparent electroencephalography correlates require symptom-domain disentangling before they can be interpreted as brain-fog related.Methods: We performed an observational secondary analysis of matched preinfection and postinfection eyes-open resting-state electroencephalography and postinfection symptom-questionnaire data from a mild SARS-CoV-2 cohort in North China. Linked electroencephalography-symptom data were available for 164 participants; the prespecified residual-analysis complete-case sample included 131 participants. The locked exposure was postinfection minus preinfection parietal-to-occipital theta phase-lag-index change from 16-channel scalp electroencephalography. Outcomes included respiratory-adjusted, total-symptom-adjusted, and severity-adjusted residual brain-fog burden, reverse residual respiratory burden, common burden, and contextual symptom-domain comparators.Findings: Brain fog was moderately correlated with respiratory burden (r = 0.371) and strongly correlated with total symptom burden (r = 0.774). The theta phase-lag-index candidate was associated with residual brain fog after respiratory adjustment (standardized beta, 0.335; 95% CI, 0.172 to 0.498; partial R2, 0.123), total-symptom adjustment (beta, 0.267; partial R2, 0.064), and severity adjustment (beta, 0.194; partial R2, 0.043). The reverse residual respiratory association was similar in magnitude (beta, -0.326; partial R2, 0.118).Interpretation: Residual brain-fog burden was partly separable from respiratory and systemic burden in association with a sensor-level theta synchrony candidate. Competing respiratory and contextual symptom-domain associations prevent exclusive, diagnostic, causal, deployable, or source-localized biomarker interpretation.

29 May 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
JSJingnan Sun
Jingnan Sun
ASAnruo Shen
Anruo Shen
YBYuda Bi
Yuda Bi
NHNora Hu
Nora Hu
YSYike Sun
Yike Sun
Physiological time-series data often violate key assumptions of classical hypothesis testing, such as independence, normality, and stationarity, limiting the reliability of widely used methods like the t-test. This study introduces the safe test framework, an e-value-based approach that enables sequential and adaptive inference while maintaining strict Type I error control under optional stopping. The study evaluates the safe test on both simulated data and five real-world physiological datasets, including EEG, ECG, EMG, PPG, and speech signals. Across simulations, the safe test demonstrates higher statistical power (0.828 vs. 0.728) and enables earlier decision-making with reduced sample sizes. In real EEG data, the safe test achieves lower empirical false negative rate (FNR) (0.62 vs. 0.69) and higher power (0.38 vs. 0.31) compared to the t-test. On ECG data, it further reduces FNR (0.53 vs. 0.64) and improves overall efficiency (0.47 vs. 0.36), while maintaining comparable empirical false positive rate (FPR). Additionally, the safe test exhibits more stable evidence accumulation and greater robustness under noise, signal degradation, and non-stationary conditions. Unlike p-values, e-values provide monotonic evidence updates, supporting reliable sequential analysis without inflating false positives. These results suggest that the safe test can provide a more stable and adaptive alternative under certain conditions, particularly in sequential and noisy data settings, making it particularly suitable for biomedical applications involving sequential data collection and noisy time-series signals.