MHANet: A hybrid attention mechanism for retinal diseases classification

PLoS One. 2021 Dec 16;16(12):e0261285. doi: 10.1371/journal.pone.0261285. eCollection 2021.

Abstract

With the increase of patients with retinopathy, retinopathy recognition has become a research hotspot. In this article, we describe the etiology and symptoms of three kinds of retinal diseases, including drusen(DRUSEN), choroidal neovascularization(CNV) and diabetic macular edema(DME). In addition, we also propose a hybrid attention mechanism to classify and recognize different types of retinopathy images. In particular, the hybrid attention mechanism proposed in this paper includes parallel spatial attention mechanism and channel attention mechanism. It can extract the key features in the channel dimension and spatial dimension of retinopathy images, and reduce the negative impact of background information on classification results. The experimental results show that the hybrid attention mechanism proposed in this paper can better assist the network to focus on extracting thr fetures of the retinopathy area and enhance the adaptability to the differences of different data sets. Finally, the hybrid attention mechanism achieved 96.5% and 99.76% classification accuracy on two public OCT data sets of retinopathy, respectively.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Choroidal Neovascularization / classification
  • Choroidal Neovascularization / diagnosis
  • Databases, Factual
  • Diabetic Retinopathy / diagnosis
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Macular Edema / classification
  • Macular Edema / diagnosis
  • Neural Networks, Computer
  • ROC Curve
  • Retina / pathology
  • Retinal Diseases / classification*
  • Retinal Diseases / diagnosis
  • Retinal Drusen / classification
  • Retinal Drusen / diagnosis
  • Retinopathy of Prematurity / classification
  • Retinopathy of Prematurity / diagnostic imaging*
  • Tomography, Optical Coherence / methods

Grants and funding

This research was funded in part by the Natural Science Foundation of Xinjiang Uygur Autonomous Region grant number 2020D01C034, Tianshan Innovation Team of Xinjiang Uygur Autonomous Region grant number 2020D14044, the National Science Foundation of China under Grant U1903213, 61771416 and 62041110, the Creative Research Groups of Higher Education of Xinjiang Uygur Autonomous Region under Grant XJEDU2017T002.