Researchers have developed RetiWave-Mamba, a novel dual-stream network designed for the early and accurate detection of retinal diseases using Optical Coherence Tomography (OCT) images. This framework integrates spatial-frequency domain learning with state-of-the-art state space models, employing Discrete Wavelet Transform to separate structural context from fine-grained details. The system achieved a state-of-the-art classification accuracy of 98.25% on the OCT-C8 dataset, demonstrating its effectiveness in identifying retinal pathologies even under noisy conditions. AI
IMPACT This research demonstrates a novel application of state space models and multi-scale analysis for improved medical image diagnostics.
RANK_REASON The cluster describes a new research paper detailing a novel deep learning network for medical image analysis. [lever_c_demoted from research: ic=1 ai=1.0]
- Attention-Guided High-Resolution Network
- Discrete Wavelet Transform
- Frequency-Adaptive Mamba Projector
- Multi-scale Contextual Localization Module
- OCT-C8 dataset
- Optical Coherence Tomography
- RetiWave-Mamba
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
