Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Automated Segmentation and Analysis of Cone Photoreceptors in Multimodal Adaptive Optics Imaging

Prajol Shrestha, Mikhail Kulyabin, Aline Sindel, Hilde R. Pedersen, Stuart Gilson, Rigmor Baraas, Andreas Maier

TL;DR

This study used advanced imaging techniques, including confocal and non-confocal split detector images from adaptive optics scanning light ophthalmoscopy (AOSLO), to analyze photoreceptors for improved accuracy and explored two U-Net-based segmentation models: StarDist for confocal and Cellpose for calculated modalities.

Abstract

Accurate detection and segmentation of cone cells in the retina are essential for diagnosing and managing retinal diseases. In this study, we used advanced imaging techniques, including confocal and non-confocal split detector images from adaptive optics scanning light ophthalmoscopy (AOSLO), to analyze photoreceptors for improved accuracy. Precise segmentation is crucial for understanding each cone cell's shape, area, and distribution. It helps to estimate the surrounding areas occupied by rods, which allows the calculation of the density of cone photoreceptors in the area of interest. In turn, density is critical for evaluating overall retinal health and functionality. We explored two U-Net-based segmentation models: StarDist for confocal and Cellpose for calculated modalities. Analyzing cone cells in images from two modalities and achieving consistent results demonstrates the study's reliability and potential for clinical application.

Automated Segmentation and Analysis of Cone Photoreceptors in Multimodal Adaptive Optics Imaging

TL;DR

This study used advanced imaging techniques, including confocal and non-confocal split detector images from adaptive optics scanning light ophthalmoscopy (AOSLO), to analyze photoreceptors for improved accuracy and explored two U-Net-based segmentation models: StarDist for confocal and Cellpose for calculated modalities.

Abstract

Accurate detection and segmentation of cone cells in the retina are essential for diagnosing and managing retinal diseases. In this study, we used advanced imaging techniques, including confocal and non-confocal split detector images from adaptive optics scanning light ophthalmoscopy (AOSLO), to analyze photoreceptors for improved accuracy. Precise segmentation is crucial for understanding each cone cell's shape, area, and distribution. It helps to estimate the surrounding areas occupied by rods, which allows the calculation of the density of cone photoreceptors in the area of interest. In turn, density is critical for evaluating overall retinal health and functionality. We explored two U-Net-based segmentation models: StarDist for confocal and Cellpose for calculated modalities. Analyzing cone cells in images from two modalities and achieving consistent results demonstrates the study's reliability and potential for clinical application.

Paper Structure

This paper contains 4 sections, 3 equations, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Materials and Methods
  3. Results
  4. Discussion

Figures (4)

  • Figure 1: Pipeline for processing multimodal AOSLO dataset, incorporating images of confocal and calculated modalities with manual and semi-automated annotation steps.
  • Figure 2: Visualization of StarDist model predictions on confocal images: cone predictions in central fovea: centers and boundaries (a), segmentation masks (b); cone predictions in parafovea: centers and boundaries (c), segmentation masks (d).
  • Figure 3: Visualization of prediction performance on the parafovea calculated image using the Cellpose 2.0 model. The white box highlights the zoomed area in the right column.
  • Figure 4: Visualizations of cone density and average cone area as a function of eccentricity: cone density fitted with asymmetric power function, volumetric cone density, average cone area computed using confocal modality (a) and calculated modality (b) images from the same participant.