Scholarlink

Purpose <p>Detecting and monitoring Microcystic Macular Edema (MME) in Optical Coherence Tomography (OCT) images is vital for early diagnosis of Diabetic Macular Edema (DME), a leading cause of blindness in developed countries. However, detecting MME remains challenging due to its fuzzy boundaries and diffuse nature. In this work, we propose a novel fully-automatic methodology based on multi-stage regional learning to successfully detect and visualize MME in OCT images.</p> Methods <p>Our work is divided into two main stages: the first stage coarsely identifies general DME accumulations in the innermost retinal layers. On the other hand, the second stage precisely detects MME within the reduced search space. These detections are then used to generate intuitive confidence maps.</p> Results <p>Our approach achieves a mean confidence of 0.9618&#xa0;±&#xa0;0.0518 per MME pixel, demonstrating consistent and strong detections. This robust methodology facilitates early diagnosis of MME, independent of clinicians’ subjectivity, and has the potential to significantly impact the quality of life of patients.</p> Conclusion <p>Our work represents a significant advancement in the automatic analysis of complex retinal pathologies. Source code is available at: <a href="https://github.com/PlacidoFranciscoLizancosVidal/Microcysts_paper_code">https://github.com/PlacidoFranciscoLizancosVidal/Microcysts_paper_code</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-Stage Learning for Intuitive Visualization of Microcystic Macular Edema in OCT Images

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年2月4日

查看全文
  • Plácido Vidal,
  • Joaquim de Moura,
  • Jorge Novo,
  • Marcos Ortega

摘要

Purpose

Detecting and monitoring Microcystic Macular Edema (MME) in Optical Coherence Tomography (OCT) images is vital for early diagnosis of Diabetic Macular Edema (DME), a leading cause of blindness in developed countries. However, detecting MME remains challenging due to its fuzzy boundaries and diffuse nature. In this work, we propose a novel fully-automatic methodology based on multi-stage regional learning to successfully detect and visualize MME in OCT images.

Methods

Our work is divided into two main stages: the first stage coarsely identifies general DME accumulations in the innermost retinal layers. On the other hand, the second stage precisely detects MME within the reduced search space. These detections are then used to generate intuitive confidence maps.

Results

Our approach achieves a mean confidence of 0.9618 ± 0.0518 per MME pixel, demonstrating consistent and strong detections. This robust methodology facilitates early diagnosis of MME, independent of clinicians’ subjectivity, and has the potential to significantly impact the quality of life of patients.

Conclusion

Our work represents a significant advancement in the automatic analysis of complex retinal pathologies. Source code is available at: https://github.com/PlacidoFranciscoLizancosVidal/Microcysts_paper_code.