AI-Driven Ocular Assessment System
ModnyCo collaborated with a client to develop groundbreaking technology enabling smart mobile devices to objectively measure the refractive state of the human eye with a single image. Leveraging deep statistical analysis and machine learning on datasets from both model and human eyes, we trained classifiers to detect eye conditions with significant accuracy. Although challenges arose in adapting the technology for human subjects, we successfully identified key variations and filed a patent application for this pioneering design. This project not only showcased ModnyCo’s expertise in optical analysis and machine learning but also marked a significant advancement in eye health diagnostics.
ModnyCo was presented with a fascinating challenge by a client: to develop a technology that enables a smart mobile device to objectively measure the refractive state of a human eye using just a single image. This project was not just about technological innovation; it was about revolutionizing the way we understand and measure eye health.
The heart of this project was creating a system that could analyse reflected light from an image and estimate the brightness of different wavelengths. This data was crucial as it helped determine the optical path length of the light rays reaching the retinal surface. Our role was to develop a classifier that could interpret this information, identifying the refractive error, axial length of the eye, or the accommodation state of the eye. This task required precision and a deep understanding of both optical physics and digital image processing.
To achieve this, ModnyCo delved into a rigorous statistical analysis, utilizing a comprehensive dataset provided by the client. This dataset was recorded using a model eye developed by the client, which served as a controlled environment for our initial investigations. Employing advanced machine learning techniques, we trained multiple classifiers, each designed to accurately detect specific eye conditions or characteristics.
Encouraged by the high accuracy of our classifiers in model eye scenarios, the client extended the experiment to include human subjects. This phase presented new challenges, primarily due to the smaller database size and variety of the human eye compared to the model eye. Despite these obstacles, our team successfully adapted the classifiers for human subjects, achieving acceptable accuracy levels. Through this process, we identified potential variations between devices and subjects, providing the client with crucial insights for future design improvements.
The success and innovation of this project culminated in ModnyCo filing a patent application for the design. This achievement was a testament to the project's novelty and potential impact on eye health diagnostics.
Throughout this project, ModnyCo demonstrated not only our technical expertise in optical analysis, machine learning, and statistical analysis but also our ability to adapt and innovate in the face of new challenges. The project with this client was a journey of discovery and achievement, paving the way for new methods in eye health diagnostics and care.
Schedule a Consultation
Connect with our team to transform your complex concepts into market-ready electronic solutions.