GCGloria C. Chiang

Gloria C. Chiang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Gloria C. Chiang has a total of 353 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Hierarchical AI classification of benign laryngeal disorders using acoustic voice features

    by Mohsen Annabestani, Samira Aghadoost, Anaïs Rameau, Olivier Elemento, Gloria Chia-Yi Chiang

    Benign laryngeal voice disorders affect nearly one in five individuals and often manifest as dysphonia, while also serving as non-invasive indicators of broader physiological dysfunction. We propose a clinically inspired hierarchical machine learning framework for automated classification of eight benign voice disorders alongside healthy controls using acoustic features extracted from short, sustained vowel phonations. Experiments were conducted on 15,132 recordings from 1,261 speakers in the Saarbrücken Voice Database, covering vowels /a/, /i/, and /u/ produced at neutral, high, low, and gliding pitches. Mimicking clinical triage workflows, the framework operates in three sequential stages: Stage 1 performs binary pathological screening by integrating convolutional neural network–derived mel-spectrogram representations with 21 interpretable acoustic features; Stage 2 groups voices into Healthy, Functional/Psychogenic, and Structural/Inflammatory categories; and Stage 3 performs fine-grained subtype classification by incorporating probabilistic outputs from preceding stages as auxiliary diagnostic features. The hierarchical framework demonstrated improved performance compared to flat multiclass classifiers and general-purpose self-supervised models (HuBERT, HeAR), suggesting the value of hierarchical decomposition for complex voice-disorder classification. A confidence-aware human-in-the-loop strategy further improved diagnostic performance by selectively referring low-confidence predictions for expert review while maintaining substantial automatic coverage, providing a practical pathway toward trustworthy clinical deployment and continual model refinement. But external validation on independent cohorts remains essential before clinical deployment. By combining deep spectral learning with interpretable voice features, the proposed framework demonstrates potential as a component of comprehensive diagnostic assessment pending clinical validation and external testing on diverse populations.

  • Liver steatosis and cognitive performance in aging: A pilot cross-sectional study

    by Mohammad Khalafi, Seyed Hani Hojjati, Xiuyuan Hugh Wang, Emily B. Tanzi, Vanessa Liao, Sonal Kumar, Neal S. Parikh, Gloria C. Chiang

  • Open Access

    Decreased cerebral metabolic rate of oxygen (CMRO 2 ) is associated with increased tau deposition in Alzheimer's disease

    by Hangwei Zhuang, Liangdong Zhou, Xiuyuan Hugh Wang, Emily B. Tanzi, Yi Li, Pascal Spincemaille, Ilhami Kovanlikaya, Thanh D. Nguyen, Yi Wang, Gloria C. Chiang

    Introduction: Transgenic Alzheimer's (AD) mouse models and post mortem studies have shown that tau pathology disrupts the mitochondrial electron transport chain and results in mitochondrial damage and synaptic dysfunction. We assessed whether magnetic resonance imaging (MRI)‐derived measures of oxygen metabolism, oxygen extraction fraction (OEF), and cerebral metabolic rate of oxygen (CMRO 2 ) were associated with tau deposition in vivo. Methods: We assessed tau using 18 F‐MK‐6240 positron emission tomography and CMRO 2 /OEF using multi‐echo gradient‐recalled echo with QQ processing, as well as pseudo‐continuous arterial spin‐labeling MRI in cognitively normal (CN) elderly and subjects with mild cognitive impairment/AD ( N = 42). Associations among imaging measures were evaluated across cortical lobes, including comparisons between regions with and without tau deposition. Results: Across all subjects, higher tau standardized uptake value ratio (SUVR) was associated with lower cerebral blood flow (CBF) and CMRO 2 . Tau SUVR was negatively associated with Montreal Cognitive Assessment scores, whereas OEF, CBF, and CMRO 2 were positively associated. Within subjects, CMRO 2 was lower in tau‐positive temporal regions in both CN and MCI/AD subjects, and CBF was lower in tau‐positive temporal regions in CN subjects. Discussion: MRI‐derived CBF and CMRO 2 were associated with regional tau deposition, particularly in the temporal lobe, which suggests they have potential as non‐invasive markers of metabolic alterations accompanying tau pathology.


  • Poor R-wave progression associates with cerebral amyloid deposition: A potential link between heart and brain

    by Stephen J. Sozio, Aditya Sahai, Mohammad Khalafi, Xiuyuan Wang, Emily Tanzi, Patrick Harvey, Thomas Maloney, Seyed H. Hojjati, Liangdong Zhou, Silky Pahlajani, et al.

    Background: Amyloid deposition is a key pathologic hallmark of Alzheimer’s disease (AD). Since cardiac amyloidosis also involves abnormal amyloid accumulation, shared proteinopathy mechanisms may underlie both conditions. Objective: We investigated whether brain amyloidosis on PET imaging is associated with electrocardiogram (EKG) findings of cardiac involvement by AD pathology. Methods: We included 191 participants (mean age 66 ± 13.0 years) in our analysis, who underwent amyloid PET imaging and EKG within one year. EKG abnormalities (including low QRS voltage, poor R-wave progression (PRWP), bundle branch block, and sinus bradycardia) were assessed for associations with brain amyloid deposition, quantified in Centiloids, and cognitive function. A subset (n = 164) also underwent 18 F-MK6240 tau PET, and the relationship between these EKG abnormalities and tau deposition was also assessed. Results: PRWP was associated with greater brain amyloid deposition (coefficient = 45 ± 12.15, p < 0.001) and worse cognition on the Clinical Dementia Rating scale (coefficient = 0.22 ± 0.098, p = 0.027). There was a trend for higher likelihood of cortical tau deposition with PRWP (odds ratio = 3.66, p = 0.05). Sinus bradycardia was associated with higher likelihood of cortical tau deposition (odds ratio = 2.81, p = 0.009). Other EKG abnormalities were not significantly associated with brain amyloid/tau deposition or cognition. Conclusions: PRWP and sinus bradycardia were found to be associated with AD pathology in the brain. These findings warrant further investigation into how cardiac electrophysiologic abnormalities may reflect a possible link between brain and cardiac pathologies or AD-associated cardiac dysfunction.


  • In Vivo [18F]MK6240 PET Reveals Atypical Patterns of Early Tau Deposition and Evidence for Primary Age-Related Tauopathy in Community-Residing Adults

    by Shama Jaswal, Bardiya Ghaderi Yazdi, Mohammad Khalafi, Sindy Ozoria, Peter Chernek, Jenseric Calimag, Xiuyuan H Wang, Qolamreza R Razlighi, Gloria C Chiang

    [18F]MK-6240, a second-generation tau PET tracer, enables in vivo detection of tau neurofibrillary tangles-one of the key pathologic hallmarks of Alzheimer's disease (AD). Building on prior work demonstrating the utility of [18F]MK-6240 for in vivo Braak staging, this study assessed the frequency of typical and atypical tau deposition patterns in a cognitively unimpaired, community-residing cohort and associations with early memory changes. One hundred seventy-three older adults (aged 69 +/- 5.7 years, 49% female) were recruited prospectively from the community to undergo [18F]MK-6240 tau PET, [18F]florbetaben amyloid PET, and clinical and neuropsychological evaluations. Images were analyzed qualitatively and quantitatively to determine amyloid and tau positivity, and participants were subsequently classified according to Braak stage. Associations with cognitive measures were assessed using linear regression models. Fifty-nine participants (34%) were found to be tau positive on PET: 32 were Braak stages I-II, 7 Braak stages III-IV, 19 Braak stages V-VI, and 1 had tau localized to the right thalamus. Notably, thirty participants (17%) were tau positive on PET but amyloid negative, suggestive of primary age-related tauopathy. Five had atypical tau deposition in cortical/subcortical regions without deposition in the medial temporal lobe, a departure from typical Braak staging, suggestive of hippocampal-sparing AD. Amyloid-positive, tau-positive participants had slightly lower cognition on the Montreal Cognitive Assessment (p=0.02) and Craft Story delayed recall (p=0.03). Amyloid-negative, tau-positive participants had lower Craft Story immediate recall scores (p=0.03). A significant proportion of older adults in the community have tau deposition in the brain, including those who are amyloid negative on PET, allowing for a critical therapeutic window before symptoms onset. Tau deposition patterns are heterogeneous and can deviate from traditional Braak staging, even in the preclinical stage, which underscores the need for an individualized, spatially-based approach to disease monitoring.

  • Open Access

    Oxygen extraction fraction is differentially associated with pathological biomarkers in Alzheimer’s disease and non-Alzheimer’s dementias

    by Arpita Misra, Yi Wang, M. Reza Taheri, Gloria C. Chiang, Junghun Cho

    Introduction: We aimed to understand the pathophysiological differences between 16 Alzheimer’s disease (AD) and 15 non-AD dementia patients by quantifying oxygen extraction fraction (OEF) in cortical (CGM) and deep gray matter (DGM) regions. Methods: To achieve this, we used a novel MRI-based OEF mapping technique, QQ, which estimates OEF from routine multi-echo gradient echo data. Multiple linear regression analyses were performed to compare the associations between OEF and white matter hyperintensities (WMH) or cognitive impairment (measured by Montreal Cognitive Assessment (MoCA) between the two groups. Results: In the AD and non-AD group, OEF showed negative associations with WMH in DGM and positive associations with MoCA in DGM and CGM. Discussion: Our study suggests that QQ is a promising tool for differentiating between AD and non-AD dementias, by revealing abnormalities in tissue oxygen usage and their relationships to microvascular changes and cognitive impairment.


  • Source-Separated Magnetic Susceptibility Reveals Postoperative Brain Changes in Newborns with Congenital Heart Disease

    by MinJung Jang, Alexey V. Dimov, Jonathan Murnick, Tarif A. Choudhury, Qinghao Liang, Kushal Kapse, Anushree Kapse, Arindam RoyChoudhury, Vivien Yap, Mary T. Donofrio, et al.

    Congenital heart disease (CHD) is a significant risk factor for neurodevelopmental impairments, yet the underlying brain alterations in the early postnatal period remain poorly understood. This prospective study measured regional magnetic susceptibilities to characterize tissue composition changes in the brain of newborns with CHD. A total of 51 newborns with CHD requiring cardiac surgery and 69 healthy controls were enrolled. Of these, 42 newborns with CHD and 65 healthy controls were included in the analysis, resulting in 29 preoperative MRI scans (postmenstrual age = 39.1 ± 1.2 weeks), 28 postoperative MRI scans (postmenstrual age = 43.0 ± 2.1 weeks), and 65 control scans (postmenstrual age = 42.8 ± 2.3 weeks). Quantitative susceptibility mapping (QSM) with source separation was performed to assess paramagnetic susceptibility and diamagnetic susceptibility. In a subset of the cohort, postoperative C-reactive protein levels were measured to assess systemic inflammation and Bayley-III test was performed at 18 months to evaluate neurodevelopment. Postoperative CHD scans demonstrated higher paramagnetic susceptibility in the thalamus and frontal lobe and higher diamagnetic susceptibility in the deep gray matter, lentiform nucleus, thalamus, and occipital lobe, compared with controls (all p < 0.05), whereas no differences were observed between preoperative CHD and control scans. Within the CHD group, single-ventricle physiology was associated with higher postoperative paramagnetic susceptibility and diamagnetic susceptibility compared with two-ventricle physiology (p < 0.05). In addition, postoperative paramagnetic susceptibility and diamagnetic susceptibility showed positive associations with postoperative C-reactive protein levels and negative associations with Bayley-III developmental scores at 18 months of age, particularly in receptive language and gross motor domains (p < 0.05). These findings suggest that susceptibility alterations emerge after cardiac surgery and are linked to poorer developmental outcomes. Source-separated QSM may serve as a sensitive biomarker of brain tissue alterations and neurodevelopmental risk in newborns with CHD.

  • Open Access

    Improving Glioblastoma Classification Using Quantitative Transport Mapping with a Synthetic Data Trained Deep Neural Network

    by Dominick J Romano, Alexandra Grace Roberts, Benjamin Weppner, Qihao Zhang, Maneesh John, Renjiu Hu, Mert Şişman, Ilhami Kovanlikaya, Gloria C Chiang, Pascal Spincemaille, et al.

    Purpose: To develop a deep neural network-based, AIF-free, perfusion estimation method (QTMnet) for improved performance on glioma classification. Methods: A globally defined arterial input function (AIF) is needed to recover perfusion parameters in the two-compartment exchange model (2CXM). We have developed Quantitative Transport Mapping (QTM) to create an AIF-independent estimation method. QTM estimation can be formulated using deep neural networks trained on synthetic DCE-MRI data (QTMnet). Here, we provide a fluid mechanics-based DCE-MRI simulation with exchange between the capillaries and extravascular extracellular space. We implemented tumor ROI generation to morphologically characterize tissue perfusion. We compared our QTMnet implementation with 2CXM on 30 glioma human subjects, 15 of which had low-grade gliomas, and 15 with high-grade glioblastomas. Results: QTMnet outperforms (best AUC: 0.973) traditional 2CXM (best AUC: 0.911) in a glioma grading task. Conclusion: The AIF-independent QTMnet estimation provides a quantitative delineation between low-grade and high-grade gliomas.

  • Open Access

    Worse visibility of deep medullary veins is associated with larger lateral ventricles but not with cortical thickness

    by Sushruth Manchineella, Henry Rusinek, Yuan Ma, Xiuyuan Hugh Wang, Surendra Maharjan, Liangdong Zhou, Tracy Butler, Yi Li, Alexus Jones, Emily Tanzi, et al.


  • Amyloid PET Z Score Quantification and Correlation with Visual Semiquantitative Grading

    by Divya Yadav, Ashley Knight-Greenfield, Joseph Moirano, Anna Nordvig, Miran W. Salgado, Moath Hamed, Michael Lin, Arindam RoyChoudhury, Sonja Blum, Samantha A. Keil, et al.

    BACKGROUND AND PURPOSE: Amyloid PET imaging plays a crucial role in the diagnosis of Alzheimer’s disease (AD) and determines eligibility for anti-amyloid therapies. While visual interpretation using Regional Cortical Tracer Uptake (RCTU) and Brain Amyloid Plaque Load (BAPL) scores remains standard in clinical practice, it is subject to inter-reader variability and may not fully depict the amyloid distribution pattern. RCTU scoring evaluates cortical tracer uptake quantified as 1: no uptake, 2: focal and 3: diffuse uptake, while BAPL provides an overall summary score of amyloid plaque load depending on highest RCTU score. Quantitative techniques such as Centiloid scale or Z-scores may assist in diagnosis, grading and treatment monitoring. This study evaluates the feasibility of a Z-score quantification generated from normative database comparison and its correlation with visual RCTU grading. MATERIALS AND METHODS: We retrospectively analyzed 100 patients who underwent [F18]-Florbetaben PET imaging between August and October 2024 for cognitive impairment. Visual interpretation was performed using RCTU scoring and overall BAPL score. Quantitative Z-scores were calculated for four cortical regions (frontal, parietal, posterior cingulate/precuneus, and lateral temporal) using a normative database. Correlation between visual and quantitative scores was assessed using Spearman’s correlation. Z-score differences among RCTU categories were evaluated with Kruskal Wallis and Mann Whitney tests. RESULTS: Among 100 patients (median age 78), 31 were amyloid negative (BAPL1), and 69 were amyloid positive (11 BAPL2, and 58 BAPL3). A total of 400 cortical regions were evaluated (143 RCTU1, 46 RCTU2, 211 RCTU3). Strong positive correlations were observed between RCTU and Z-scores in all regions (ρ = 0.78–0.88, p < 0.0001). The regional Z-scores showed significant differences between RCTU1 and RCTU2, RCTU1 and RCTU3, as well as between RCTU2 and RCTU3 across all four regions (p<0.05 for all comparisons. Pooled regional analysis also showed statistically significant differences in Z-scores between all RCTU groups (p < 0.0001). CONCLUSIONS: Quantitative regional Z-scores derived from amyloid PET imaging demonstrate a strong correlation with visual assessment (RCTU score), validating their feasibility in clinical interpretation. These findings support the integration of quantitative tools into routine practice to enhance diagnostic confidence, reduce reader variability, and monitoring for amyloid-targeted therapies. ABBREVIATIONS: AD = Alzheimer’s Disease; RCTU = Regional Cortical Tracer Uptake; BAPL = Brain Amyloid Plaque Load; PiB = Pittsburgh Compound B; SUVR = Standardized Uptake Value Ratio; SPM = Statistical Parametric Mapping; SSP = Stereotactic Surface Projections.