Interferometer eye image classification for dry eye categorization using phylogenetic diversity indexes for texture...

Area of Science:

• Ophthalmology and Biomedical Imaging
• Computational Biology and Bioinformatics

Background:

• Dry eye syndrome significantly impacts quality of life, necessitating accurate diagnostic methods.
• Current diagnosis relies on manual classification of tear film interferometry images, which is subjective.
• Automated systems are needed to support expert diagnosis of tear film instability.

Purpose of the Study:

• To develop an automated method for classifying tear film lipid layers.
• To utilize phylogenetic diversity indexes for feature extraction in tear film images.
• To evaluate the efficacy of various machine learning classifiers for this task.

Main Methods:

• Acquisition of the VOPTICAL_GCU image dataset.
• Segmentation of the region of interest within the images.
• Feature extraction using phylogenetic diversity indexes.
• Classification using Support Vector Machines, Random Forest, Naive Bayes, Multilayer Perceptron, Random Tree, and RBFNetwork algorithms.
• Validation of the classification results.

Main Results:

• The Random Forest classifier achieved the highest accuracy, exceeding 97%.
• Performance metrics included a standard deviation of 0.51%, AUC of 0.99, Kappa index of 0.96, and F-Measure of 0.97.
• The proposed method demonstrated high efficiency in classifying tear film lipid layers.

Conclusions:

• Phylogenetic diversity indexes are effective for feature extraction in tear film lipid layer classification.
• The developed automated system provides efficient and accurate diagnostic support for dry eye syndrome.
• This approach offers a promising solution for objective tear film analysis.