The primary goal of the proposed research program is to develop computer software tools with embedded artificial intelligence (AI) that can perform instantaneous, automated analysis and clinical interpretation of wavefront error measurements of the human eye and cornea. Secondary goals are to improve the overall design of oculometric data visualization tools, provide information that will help to establish clinical and scientific standards for ocular measurements and procedures, and improve our understanding of the fundamental relationship between optical performance and visual performance. We hypothesize that a) AI-based algorithms will detect complex patterns of wavefront errors; b) these patterns are specific to and significantly correlated with certain diseases and disorders; and c) AI-based interpretation of complex data will be superior to that performed by expert humans, who are the gold standard for interpreting clinical data. Specifically, we will (1) develop, train, and test AI-based algorithms (Bayesian and neural networks) to interpret the significance of complex wavefront error data obtained retrospectively from examination records of patients with various ocular diseases, disorders, or surgical interventions, as well as normal eyes; (2) simulate wavefront error data using computer models based on statistical distributions of actual ocular aberrations from patient population samples for the purpose of investigating the importance of individual higher order aberrations to retinal image formation and potential visual performance, as well as to generate new data that will enhance the overall AI training and testing process, and (3) establish standard methods to acquire and analyze wavefront error data. AI-based tools will assist vision scientists to efficiently develop study databases and analyze aberration data. Clinicians will diagnose patients faster, more accurately, and with a greater degree of confidence. For patients, refractive surgery outcomes will be more predictable, and they will benefit from earlier detection of diseases such as cataracts and corneal ectasias.
| Smolek, Michael K (2012) Method for expressing clinical and statistical significance of ocular and corneal wave front error aberrations. Cornea 31:212-21 |
| Smolek, Michael K; Klyce, Stephen D (2007) Absolute color scale for improved diagnostics with wavefront error mapping. Ophthalmology 114:2022-30 |
| Smolek, Michael K; Klyce, Stephen D (2005) Goodness-of-prediction of Zernike polynomial fitting to corneal surfaces. J Cataract Refract Surg 31:2350-5 |
| Klyce, Stephen D; Karon, Michael D; Smolek, Michael K (2004) Advantages and disadvantages of the Zernike expansion for representing wave aberration of the normal and aberrated eye. J Refract Surg 20:S537-41 |
