Acquire continuous and precise score for fundus image quality assessment: FTHNet and FQS dataset

TL;DR

This paper introduces a new dataset and a deep learning model for assessing the quality of retinal fundus images with high accuracy and efficiency.

Contribution

The novel FTHNet model and the FIQA dataset with continuous quality scores improve fundus image quality assessment for clinical use.

Findings

FTHNet achieves high correlation scores (Pearson 0.9423, Spearman 0.9488) in predicting fundus image quality.

FTHNet outperforms existing methods with fewer parameters and lower computational complexity.

The new dataset includes 2,246 images with continuous quality scores and three-level categories.

Abstract

The retinal fundus images are extensively utilized in diagnosis, and their quality may affect diagnostic results. However, due to limitations in the datasets and algorithms, current fundus image quality assessment (FIQA) methods often lack the granularity required to meet clinical demands. To address these limitations, we introduce a new benchmark FIQA dataset, Fundus Quality Score, which contains 2,246 images annotated with continuous mean opinion scores ranging from 0 to 100 and three-level quality categories. Meanwhile, we also design a novel FIQA Transformer-based Hypernetwork (FTHNet). The FTHNet can treat FIQA as a regression task to predict the continuous MOS, diverging from common classification-based approaches. Results on our dataset show that FTHNet predicts quality scores, achieving a Pearson Linear Correlation Coefficient of 0.9423 and a Spearman Rank Correlation Coefficient of 0.9488, significantly outperforming compared methods while utilizing fewer parameters and lower computational complexity. Furthermore, model deployment experiments demonstrate its potential for use in automated medical image quality control workflows. We have released the code and dataset to facilitate future research in this field.

Figures