CATRA: interactive measuring and modeling of cataracts

Vitor F. Pamplona, Erick Baptista Passos, Jan Zizka, Manuel M. Oliveira, Everett Lawson, Esteban Walter Gonzalez Clua, Ramesh RaskarACM Trans. Graphics 2011 · DOI

TL;DR

This paper introduces CATRA, a novel interactive system that uses a light-field display and parallax barriers to measure the perceptual impact of forward scattering from cataracts in the human eye. Unlike traditional methods, CATRA quantifies opacity, attenuation, contrast, and point-spread functions (PSFs) of cataracts, enabling a personalized simulation of a cataract-affected view for better understanding and diagnosis.

What Problem Does This Paper Address?

Cataracts are a leading cause of avoidable blindness, characterized by clouding of the eye's crystalline lens that scatters light. Current methods for detecting and assessing cataracts often rely on highly trained clinicians and subjective evaluations, or expensive, complex equipment. There's a need for accessible, user-friendly tools that can quantitatively measure the impact of cataracts on vision, particularly the forward scattering that blurs vision and reduces contrast, and to provide a way for individuals to understand how their vision is affected.

Methods

CATRA employs a unique interactive approach using a dual-stacked LCD display and parallax barriers to create collimated light beams that scan the crystalline lens. The system works in several steps:

  • Pupil Size Measurement: A simplified ray diagram is used to estimate pupil size, which is crucial for scaling measurements.
  • Cataract Presence Screening: The system automatically scans the lens, and users provide feedback (e.g., by noticing blinking or fading patterns) to detect the presence of cataracts.
  • Opacity and Attenuation Mapping: If cataracts are detected, users interactively mark faded regions (opacity map) and then match the brightness of clear and scattered light paths to create an attenuation map, indicating the intensity of scattering.
  • Point-Spread Function (PSF) Measurement: For more detailed analysis, users match projected patterns to estimate the PSF of the cataract. If the scattering is too strong for direct PSF matching, a contrast-sensitivity test is used as a proxy.
  • Simulation of Subject's View: The collected opacity, attenuation, and PSF maps are used to render a simulation of how the individual's vision is affected by their cataracts, incorporating depth-dependent scattering.

The system is designed to be compact, affordable, and easy to use, replacing mechanical scanning with on-screen pattern movement.

Datasets

The paper does not explicitly mention large, pre-existing datasets for training or evaluation. Instead, it focuses on:

  • User Studies: Evaluations were conducted with 18 subjects using a cell-phone prototype and with 5 early cataract-affected volunteers.
  • Camera-based Simulations: The authors used a DSLR camera with diffusers to simulate cataracts and compare their system's measurements (attenuation and PSF maps) against the simulated visual effects. This served as a controlled evaluation method.

Key Insights and Ideas

  • Interactive Perceptual Matching: The core innovation is using user perception (matching brightness, identifying patterns) as a feedback mechanism to quantify complex optical properties like scattering and PSF. This makes the system accessible to non-experts.
  • Forward Scattering Focus: By focusing on forward scattering, CATRA addresses a key aspect of cataract vision impairment that is often less emphasized than backscattering in traditional clinical assessments.
  • Personalized Vision Simulation: The ability to reconstruct an individual's cataract-affected view is a significant step towards bridging the gap between clinical measurements and the patient's lived experience, fostering better communication and understanding.
  • Compact and Affordable Design: The use of off-the-shelf displays and simple optical components makes the technology potentially deployable in low-resource settings.

Key Results

  • Quantitative Maps: CATRA successfully generates quantitative maps for opacity, attenuation, contrast, and sub-aperture PSFs.
  • Accuracy and Repeatability:
    • Attenuation map accuracy showed an average error of 1.03% ± 4.20%.
  • Contrast sensitivity repeatability had an average error of 0.03 ± 0.03.
  • PSF map repeatability was 0.03 ± 0.13 in normalized scale.
  • Size measurements for simulated cataracts showed good agreement with ground truth (e.g., 34mm² ± 4.27 for an 80° diffuser).
  • Repeatability for diffuser centroids was excellent (0.01mm ± 0.01).
  • User Evaluation: 18 subjects tested the cell-phone prototype. While some found the alignment task challenging, the overall response was positive, with many fascinated by their opacity maps. 14 healthy eyes were scanned with no cataracts detected.
  • Simulation Fidelity: The rendered simulations using measured PSFs showed visual effects similar to those seen with actual simulated cataracts and photographs.

Limitations and Future Work

  • User Participation Dependency: The system relies on active user participation, which can be a limitation if the user is unable to follow instructions or if the cataract is uniformly scattering, making it difficult to discern patterns.
  • Advanced Cataracts: Uniformly scattering cataracts may obscure the ability to mark scattering regions for opacity maps.
  • Retinal Diseases: The presence of other retinal diseases could affect the results.
  • Refractive Aberrations: While the system is sensitive to refractive errors, these can be accounted for by pre-warping light rays.
  • Display Limitations: Dual-stacked LCDs have inherent limitations like reduced brightness and viewing zones.
  • Future Work: The authors suggest exploring hierarchical procedures for scanning, developing more robust alignment tasks, and further clinical studies to validate the impact of refractive aberrations. They also highlight the potential for this technology to be used as a screening tool for other visual impairments and to address the lack of glare disability tests in current licensing procedures.

Summary updated

BibTeX
@article{pamplona2011catra,
  author = {Vitor F. Pamplona and Erick Baptista Passos and Jan Zizka and Manuel M. Oliveira and Everett Lawson and Esteban Walter Gonzalez Clua and Ramesh Raskar},
  title = {{CATRA: interactive measuring and modeling of cataracts}},
  journal = {ACM Trans. Graphics},
  year = {2011},
  doi = {10.1145/2010324.1964942},
}
Figure