Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Dual-Modality Computational Ophthalmic Imaging with Deep Learning and Coaxial Optical Design

Boyuan Peng, Jiaju Chen, Yiwei Zhang, Cuiyi Peng, Junyang Li, Jiaming Deng, Peiwu Qin

TL;DR

This work tackles the need for accessible ophthalmic screening by introducing a compact dual-modality device that combines fundus photography and refractive error assessment within a single coaxial optical design. It leverages wavelength-based separation and a Dense-U-Net for pupil segmentation to enable automated eye alignment and focusing, supplemented by a unidirectional connected component analysis for pupil coordinate extraction. Quantitatively, the system achieves precise pupil localization and robust refractive estimation, while image quality iscurrently limited by commercial lenses and uncorrected high-order aberrations. The approach shows promise for scalable, AI-assisted screening in community settings, with potential impact on early detection of myopia progression and retinal diseases.

Abstract

The growing burden of myopia and retinal diseases necessitates more accessible and efficient eye screening solutions. This study presents a compact, dual-function optical device that integrates fundus photography and refractive error detection into a unified platform. The system features a coaxial optical design using dichroic mirrors to separate wavelength-dependent imaging paths, enabling simultaneous alignment of fundus and refraction modules. A Dense-U-Net-based algorithm with customized loss functions is employed for accurate pupil segmentation, facilitating automated alignment and focusing. Experimental evaluations demonstrate the system's capability to achieve high-precision pupil localization (EDE = 2.8 px, mIoU = 0.931) and reliable refractive estimation with a mean absolute error below 5%. Despite limitations due to commercial lens components, the proposed framework offers a promising solution for rapid, intelligent, and scalable ophthalmic screening, particularly suitable for community health settings.

Dual-Modality Computational Ophthalmic Imaging with Deep Learning and Coaxial Optical Design

TL;DR

This work tackles the need for accessible ophthalmic screening by introducing a compact dual-modality device that combines fundus photography and refractive error assessment within a single coaxial optical design. It leverages wavelength-based separation and a Dense-U-Net for pupil segmentation to enable automated eye alignment and focusing, supplemented by a unidirectional connected component analysis for pupil coordinate extraction. Quantitatively, the system achieves precise pupil localization and robust refractive estimation, while image quality iscurrently limited by commercial lenses and uncorrected high-order aberrations. The approach shows promise for scalable, AI-assisted screening in community settings, with potential impact on early detection of myopia progression and retinal diseases.

Abstract

The growing burden of myopia and retinal diseases necessitates more accessible and efficient eye screening solutions. This study presents a compact, dual-function optical device that integrates fundus photography and refractive error detection into a unified platform. The system features a coaxial optical design using dichroic mirrors to separate wavelength-dependent imaging paths, enabling simultaneous alignment of fundus and refraction modules. A Dense-U-Net-based algorithm with customized loss functions is employed for accurate pupil segmentation, facilitating automated alignment and focusing. Experimental evaluations demonstrate the system's capability to achieve high-precision pupil localization (EDE = 2.8 px, mIoU = 0.931) and reliable refractive estimation with a mean absolute error below 5%. Despite limitations due to commercial lens components, the proposed framework offers a promising solution for rapid, intelligent, and scalable ophthalmic screening, particularly suitable for community health settings.

Paper Structure

This paper contains 13 sections, 18 equations, 5 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Methods
  3. Optical Design of Fundus Photography and Refraction Systems.
  4. Integrated Optical Design for Combined Fundus Imaging and Refraction.
  5. U-Net-Based Pupil Localization Method.
  6. Pupil Coordinate Extraction Method Based on Unidirectional Connected Component Analysis.
  7. Results
  8. Ocular surface structure segmentation.
  9. Pupil coordinate parameter extraction.
  10. Fundus photography image quality assessment.
  11. Fundus photography image quality assessment.
  12. Discussion
  13. Conclusion

Figures (5)

  • Figure 1: Construction of system; (a) Fundus imaging; (b) Refraction optical paths
  • Figure 2: Integrated Optical Path for Dual Examination; (a) Coupled optical path; (b) Photomechanical structure
  • Figure 3: Extraction of pupil coordinates; (a) Hough Circle Transform; (b) Canny; (c) Unidirectional closed connected domain extraction strategy; (d) Pupil location algorithm results
  • Figure 4: Pupil image processing; (a) Preprocessing; (b) Segmentation results of validation set and model eyes
  • Figure 5: Comparison of imaging using 2-in-1 devices and commercial devices