Automatic classification of the interferential tear film lipid layer using colour texture analysis

• Published in: Computer methods and programs in biomedicine Vol. 111; no. 1; pp. 93 - 103
• Main Authors: Remeseiro, B., Penas, M., Barreira, N., Mosquera, A., Novo, J., García-Resúa, C.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ireland Ltd 01.07.2013; Elsevier
• Subjects: Adult; Applied sciences; Artificial intelligence; Biological and medical sciences; Color; Computer science; control theory; systems; Data processing. List processing. Character string processing; Databases, Factual; Exact sciences and technology; Guillon categories; Humans; Internal Medicine; Investigative techniques, diagnostic techniques (general aspects); Lipids - chemistry; Lipids - classification; Machine learning; Markov Chains; Medical sciences; Memory organisation. Data processing; Microscopy, Interference - statistics & numerical data; Ophthalmology; Optical Phenomena; Other; Pathology. Cytology. Biochemistry. Spectrometry. Miscellaneous investigative techniques; Pattern recognition. Digital image processing. Computational geometry; Principal Component Analysis; Software; Support Vector Machine; Tear film lipid layer; Tears - chemistry; Texture analysis; Wavelet Analysis; Young Adult; Tear film lipid layer; Guillon categories; Texture analysis; Machine learning; Principal component analysis; Automatic classification; Dimensionality; Tear; Pattern recognition; Texture; Color image; Thickness; Selection criterion; Ophthalmology; Artificial intelligence; Pattern extraction
• ISSN: 0169-2607; 1872-7565; 1872-7565
• DOI: 10.1016/j.cmpb.2013.04.007

The tear film lipid layer is heterogeneous among the population. Its classification depends on its thickness and can be done using the interference pattern categories proposed by Guillon. This papers presents an exhaustive study about the characterisation of the interference phenomena as a texture pattern, using different feature extraction methods in different colour spaces. These methods are first analysed individually and then combined to achieve the best results possible. The principal component analysis (PCA) technique has also been tested to reduce the dimensionality of the feature vectors. The proposed methodologies have been tested on a dataset composed of 105 images from healthy subjects, with a classification rate of over 95% in some cases.