Artificial Intelligence in Glaucoma: Advances in Diagnosis, Progression Forecasting, and Surgical Outcome Prediction

International Journal of Molecular Sciences, Journal Year: 2025, Volume and Issue: 26(10), P. 4473 - 4473

Published: May 8, 2025

Glaucoma is a leading cause of irreversible blindness, with challenges persisting in early diagnosis, disease progression, and surgical outcome prediction. Recent advances artificial intelligence have enabled significant progress by extracting clinically relevant patterns from structural, functional, molecular data. This review outlines the current applications glaucoma care, including detection using fundus photography OCT progression prediction deep learning architectures such as convolutional neural networks, recurrent transformer models, generative adversarial autoencoders. Surgical forecasting has been enhanced through multimodal models that integrate electronic health records imaging We also highlight emerging AI omics analysis, transcriptomics metabolomics, for biomarker discovery individualized risk stratification. Despite these advances, key remain interpretability, integration heterogeneous data, lack personalized timing guidance. Future work should focus on transparent, generalizable, supported large, well-curated datasets, to advance precision medicine glaucoma.

Language: Английский

HistoGWAS: An AI-enabled Framework for Automated Genetic Analysis of Tissue Phenotypes in Histology Cohorts

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: June 12, 2024

Abstract Understanding how genetic variation affects tissue structure and function is crucial for deciphering disease mechanisms, yet comprehensive methods analysis of histology are lacking. We address this gap with HistoGWAS, a framework integrating AI tools representation learning image generation fast variance component models to enable scalable interpretable genome-wide association studies histological traits. HistoGWAS employs foundation automated trait characterization generative visually interpret the influences on these Applied eleven types from GTEx cohort, identifies four significant loci, which we linked specific gene expression changes. A power confirms effectiveness in analyses large-scale data, underscoring its potential transform imaging studies.

Language: Английский