UNEATLANTICO takes part in a study on the detection of retinitis pigmentosa through deep learning

25 Mar 2024
UNEATLANTICO takes part in a study on the detection of retinitis pigmentosa through deep learning

A researcher at the Higher Polytechnic School of the Universidad Europea del Atlántico (European University of the Atlantic, UNEATLANTICO) is part of a study that proposes a new method for detecting retinitis pigmentosa through deep learning.

The retina, one of the most active tissues in the human body, can suffer alterations in its structure due to various diseases. Early detection of these alterations is important for diagnosis and treatment. Fundus imaging and optical coherence tomography (OCT) are two methods for examining eye conditions, such as retinitis pigmentosa (RP), diabetic retinopathy, and macular degeneration.

RP is a group of inherited disorders of the retina that result in degeneration of the photoreceptor cells. This can cause a gradual loss of vision and currently has no cure. Initial signs of RP usually include loss of night vision and loss of vision in mid-peripheral areas, which may progress to complete loss of vision.

Both OCT and fundus imaging are useful for the analysis of RP. However, sometimes the visual analysis performed by physicians is not optimal due to various factors, such as lack of experience and image quality. Therefore, artificial intelligence-based algorithms are being implemented to improve the detection and diagnosis of eye diseases. An example of this is the specialized model called Se (Res-Net), which compresses image information to better detect and analyze signs of RP in the retina.

In this study, a Se-ResNet-based neural network architecture is proposed, for accurate and automatic detection of RP in color fundus images. This model uses SE blocks and residual learning to improve image representation and segmentation capabilities. The results indicate that the proposed model is sensitive and specific in the detection of RP, which makes it a useful tool for physicians in the evaluation of disease progression and severity, allowing earlier diagnosis and better use of available resources.

If you want to know more about this fascinating study, click here

For further research, check the UNEATLANTICO repository