Wang Mengyu

Sun Jessica A, Johnson Grace E, Lee Kyoung Min, Zhao Yan, Saini Chhavi, Wang Mengyu et al.

This study found POAG patients have greater inter-eye nerve fiber layer asymmetry, but similar microvascular asymmetry to controls, suggesting systemic factors, like hypertension treatment, may influence glaucoma's vascular component.

Zhang Yapei, Shi Min, Chung In Young, Liebman Daniel L, Barna Laura E, Pasquale Louis R et al.

LLaMA 2-70B shows promise for patient-friendly glaucoma summaries, achieving 66-70% accuracy. Better for suspects, it excels in targeted questions (80-91% accuracy). LLMs could enhance patient communication and health literacy in ophthalmology.

Majid Iyad, Zebardast Nazlee, Wang Mengyu

AI translated underutilized SLO images into synthetic color fundus photos. Using these synthetic images with real ones significantly improved deep learning glaucoma detection, enhancing screening accuracy.

Lokhande Anagha, Aziz Kanza, Fujita Asahi, Pasquale Louis R, Shen Lucy Q, Friedman David S et al.

This study found most glaucoma surgeries, especially complex ones like trabeculectomy, occur in urban centers, often requiring significant patient travel. This highlights disparities in surgical access, particularly for rural patients.

Katuru Abhilash, Chung In Young, Majid Iyad, Shen Lucy Q, Wang Mengyu

Deep learning with OCT scans detected glaucoma more accurately than disc photos, showing OCT's objective value. This supports OCT-based AI for glaucoma, but highlights demographic bias in current models.

Sekimitsu Sayuri, Selvan Kavin, Zhao Yan, Hashemabad Saber Kazeminasab, Eslami Mohammad, Wang Mengyu et al.

This study found genetic risk for glaucoma subtypes correlates with specific visual field defects; higher POAG/NTG genetic risk linked to paracentral loss, while HTG genetic risk was associated with total loss, impacting diagnosis and prognosis.

McCalla Aliah, Wang Mengyu, Eslami Mohammad, Kazeminasab Saber, Luo Yan, Rana Hannah et al.

This study found that cup-to-disc ratio (CDR) has ceiling effects and poorly detects early glaucoma damage, especially in screening. Clinicians should interpret CDR values cautiously.

Pang Yueyin, Bang Ji Won, Kasi Anisha, Li Jeremy, Parra Carlos, Fieremans Els et al.

This study linked specific brain microstructural changes and metabolic deficits (dMRI, MRS) in glaucoma patients to distinct visual field loss patterns, highlighting neuroimaging's role in understanding glaucomatous neurodegeneration.

Sun Jessica A, Yuan Melissa, Johnson Grace E, Pasquale Louis R, Boland Michael V, Friedman David S et al.

This study found PACG patients had less structural damage but similar functional loss, with PACG showing inferior field loss and POAG superior, aiding differentiated monitoring.

Shi Min, Sun Jessica A, Lokhande Anagha, Tian Yu, Luo Yan, Elze Tobias et al.

Deep learning corrected RNFLT artifacts, significantly improving visual field and glaucoma progression prediction, enhancing the clinical utility of otherwise compromised OCT scans.

Singh Rishabh, Rauscher Franziska G, Li Yangjiani, Eslami Mohammad, Kazeminasab Saber, Zebardast Nazlee et al.

This study found that OCT's normative percentiles of retinal nerve fiber layer thickness did not improve prediction of glaucomatous visual field loss compared to raw thickness measurements, challenging current clinical assumptions.

Yousefi Siamak, Huang Xiaoqin, Poursoroush Asma, Majoor Julek, Lemij Hans, Vermeer Koen et al.

An AI system objectively staged glaucoma severity using RNFL thickness, finding optimal thresholds for normal, early, moderate, and advanced loss, offering a consistent clinical tool.

Eslami Mohammad, Kazeminasab Saber, Sharma Vishal, Li Yangjiani, Fazli Mojtaba, Wang Mengyu et al.

This Python package (PyVisualFields) was developed to enable visual field analysis in Python, bridging a gap for AI-driven glaucoma research and facilitating clinical application development.

Eslami Mohammad, Kim Julia A, Zhang Miao, Boland Michael V, Wang Mengyu, Chang Dolly S et al.

Deep learning models for glaucoma visual field prediction underpredict worsening vision loss, performing poorly in progressive cases, limiting their current clinical utility for identifying patients needing intervention.

Kang Jae H, Wang Mengyu, Frueh Lisa, Rosner Bernard, Wiggs Janey L, Elze Tobias et al.

This study found Black individuals have a significantly higher risk of both early central and advanced visual field loss patterns in glaucoma compared to non-Hispanic Whites, highlighting racial disparities in disease progression.

Huang Xiaoqin, Saki Fatemeh, Wang Mengyu, Elze Tobias, Boland Michael V, Pasquale Louis R et al.

AI developed a glaucoma staging system using visual fields, finding four severity levels based on specific MD thresholds. This objective, easy-to-use system aids clinical practice and research.

Halawa Omar A, Jin Qingying, Pasquale Louis R, Kang Jae H, Lorch Alice C, Sobrin Lucia et al.

This study found Black, Asian, and Hispanic patients had worse baseline glaucoma. Black patients also experienced less frequent testing and faster progression, highlighting disparities needing targeted interventions.

Huang Xiaoqin, Sun Jian, Majoor Juleke, Vermeer Koenraad Arndt, Lemij Hans, Elze Tobias et al.

AI accurately estimated visual field damage (MD) from OCT RNFL thickness better than other models, showing potential to objectively assess glaucoma severity and reduce reliance on subjective visual field tests.

Shoji Marissa K, Cousins Clara C, Saini Chhavi, Nascimento E Silva Rafaella, Wang Mengyu, Brauner Stacey C et al.

This study found primary open-angle glaucoma patients have reduced microvascular blood flow in both the optic nerve and nailfold capillaries, suggesting systemic microvascular dysfunction is relevant to glaucoma treatment.

Wang Mengyu, Shen Lucy Q, Pasquale Louis R, Wang Hui, Li Dian, Choi Eun Young et al.

AI identified 16 distinct RNFLT patterns in glaucoma. These patterns better predicted visual field loss than traditional methods, potentially improving clinical assessment of RNFLT damage.

Baniasadi Neda, Rauscher Franziska G, Li Dian, Wang Mengyu, Choi Eun Young, Wang Hui et al.

This study mapped normal interocular RNFL thickness differences across 768 locations, finding significant, location-specific asymmetries influenced by age and scan radius, which can improve early glaucoma detection.

Yousefi Siamak, Elze Tobias, Pasquale Louis R, Saeedi Osamah, Wang Mengyu, Shen Lucy Q et al.

An AI dashboard effectively mapped diverse glaucoma visual field loss patterns, showing 77% sensitivity for progression. This offers clinicians a user-friendly tool to monitor disease severity and progression.

Wang Mengyu, Tichelaar Jorryt, Pasquale Louis R, Shen Lucy Q, Boland Michael V, Wellik Sarah R et al.

AI identified 14 central visual field patterns in end-stage glaucoma. Key finding: initial central loss is often nasal, with one specific nasal pattern prone to total blindness, aiding prognosis and management.

Wang Mengyu, Shen Lucy Q, Pasquale Louis R, Boland Michael V, Wellik Sarah R, De Moraes Carlos Gustavo et al.

AI classified central visual field patterns in glaucoma, identifying 17 distinct loss types. This improved prediction of future central VF worsening, offering better prognostic insights for patients.

Li Dian, Rauscher Franziska G, Choi Eun Young, Wang Mengyu, Baniasadi Neda, Wirkner Kerstin et al.

This study found significant, location-specific sex differences in RNFL thickness, with females generally thicker but males thicker superiorly. Incorporating sex into RNFL norms could improve glaucoma diagnosis.