Related Experiment Video
Updated: Aug 5, 2026

LipidUNet-Machine Learning-Based Method of Characterization and Quantification of Lipid Deposits Using iPSC-Derived Retinal Pigment Epithelium
Published on: July 28, 2023
Modeling Retinal Damage for Threshold Prediction and Probabilistic Risk Assessment
Nico Heussner1, Adrian Azarian2
1Fraunhofer-Institute of Optronics, System Technologies and Image Exploitation IOSB; nico.heussner@iosb.fraunhofer.de.
Abstract:
With the increasing number of laser applications in medicine, defense, and technology, both intentional and accidental exposure of the human eye to laser sources has become a major concern. Model-based prediction of retinal damage thresholds could enable more scenario-specific laser safety assessment, especially for laser parameters not covered by experimental data. Ideally, such models would allow the calculation of ED50 (effective dose at which the likelihood of damage is 50%) values based -among other factors - on wavelength, pulse duration, and spot shape. This requires a detailed understanding and modeling of all damage regimes to reflect the dependency between the key parameters and the dominant damage mechanism. This work discusses the status of this approach (validated here for the thermal regime, or simply "in the thermal damage regime"); critical aspects that could block its success are highlighted, and potential benefits are outlined. These range from improved accuracy of laser exposure limits in eye safety standards to optimized dosimetry in retinal laser surgery to probabilistic risk assessment for the use of lasers in outdoor environments.
