Researchers have developed advanced AI frameworks for glaucoma diagnosis, aiming to improve upon opaque deep-learning models. GlaKG utilizes a knowledge graph to provide traceable reasoning by integrating biomarkers, clinical rules, and image features, achieving high accuracy in classification and risk stratification. GlaBoost employs a multimodal gradient boosting approach, combining fundus image embeddings, text-based assessments, and structured biomarkers for enhanced, interpretable predictions. Another framework uses a Vision Transformer (ViT) with a stacking ensemble to process fundus images and clinical data, offering strong performance in both sample-wise and patient-wise detection and a deployed web platform for screening. AI
IMPACT These frameworks offer more interpretable and accurate AI-driven diagnostic tools, potentially improving patient outcomes in ophthalmology.
RANK_REASON The cluster consists of three research papers published on arXiv detailing novel AI frameworks for medical diagnosis.
- alphaXiv
- arXiv
- biomarker
- CatalyzeX
- clinical data
- DagsHub
- GlaBoost
- GlaKG
- glaucoma
- Gotit.pub
- Hugging Face
- ScienceCast
- Vision Transformer
- ViT
- XGBoost
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →
