Optics

The group is currently working on a line of research based on the analysis of the preocular tear film in eye photographies and videos. This line of research is directed to the diagnosis of various tear dysfunction problems, like the «dry eye syndrome».

Collaborations

Optometry Group.
Department of Applied Physics.

Department of Life Sciences.

Centro de Física.
Universidade do Minho

Working fields

Tear film lipid layer classification using Tearscope images

We have developed a methodology to classify the tear film lipid layer patterns using the Tearscope plus as the instrument to acquire tear film images. The proposed methodology works on several stages to detect the region of interest of an input image, extract its descriptor based on color and texture information, and classify it into one of the five categories: open meshwork, close meshwork, wave, amorphous and color fringe. The method has been tested on several images from each tear type with maximum accuracy over 95%. Additionally, we have developed another methodology to create tear film maps, which illustrates the spatial heterogeneity of the lipid layer and allows to detect multiple patterns per patient.

Tear film lipid layer classification using Doane images

We have developed a methodology to classify the tear film lipid layer using the Doane interferometer as the instrument to acquire tear film images. The proposed methodology works on several stages to detect the region of interest of an input image, extract its descriptor based on color and texture information, and classify it into one of the five categories: strong fringes, coalescing strong fringes, fine fringes, coalescing fine fringes and debris. The method has been tested on several images from each tear type with maximum accuracy over 93%.

Tear film break-up time test

The Break-Up Time (BUT) is a clinical test which consists of measuring the time that the tear film remains stable without blinking. We have developed an automatic methodology to perform this test. The proposed methodology locates the different measurement areas from a tear film video, extracts the region of interest and performs the BUT test in each measurement area. The method has been tested on 18 tear film videos annotated by 4 different optometrists achieving results in the same range as between the experts themselves.

Public databases

Tearscope images

These datasets contain images of the preocular tear film over different illumination conditions. All the images have a spatial resolution of 1024×768 pixels and have been acquired with the Tearscope Plus (Keeler, Windsor, UK). All these datasets have been acquired and annotated by optometrists from the Optometry Service of the University of Santiago de Compostela (Spain) and, in the last case, also by optometrists from the Physics Center of the University of Minho (Portugal). The filenames of the images indicate the predominant interference pattern: CO is color fringe, AM is amorphous, FL is wave, MA is open meshwork, and MC is closed meshwork. If you are interested in using the VOPTICAL datasets, please send an email to: bremeseiro@uniovi.es and you will receive an authentication password to access the datasets. This is for statistical purposes only, no private data or fee is required. If you use this image set for your work, please include a reference to the database in it.
  • The VOPTICAL_I1 Database distribution [Cite] contains 105 images of the preocular tear film taken over optimum illumination conditions, and acquired from health subjects with dark eyes and aged from 19 to 33 years. The dataset includes 29 open meshwork, 29 closed meshwork, 25 wave and 22 color fringe images.
    The VOPTICAL_I1 Database processed contains 588 features extracted from the 105 images using the Lab color space and the co-occurrence features technique for color and texture analysis, respectively. More information can be found in «A Methodology for improving tear film lipid layer classification» [DL]. Note that the correspondence between labels of classes and interference patterns is: 1 for color fringe, 2 for wave, 3 for open meshwork, and 4 for closed meshwork.
  • The VOPTICAL_I1-v2 Database distribution [Cite] is the updated version of the VOPTICAL_I1 dataset. It contains 128 images of the preocular tear film taken over optimum illumination conditions, and acquired from healthy subjects with dark eyes and aged from 19 to 33 years. The dataset includes 29 open meshwork, 29 closed meshwork, 25 wave, 23 amorphous and 22 color fringe images.
  • The VOPTICAL_L Database distribution [Cite] contains 108 images of the preocular tear film taken over optimum illumination conditions, and acquired from healthy subjects with light eyes and aged from 19 to 33 years. The dataset includes 30 open meshwork, 28 closed meshwork, 27 wave and 23 color fringe images.
  • The VOPTICAL_Is Database distribution [Cite] contains 406 images of the preocular tear film taken over four different illuminations, and acquired from healthy subjects with dark eyes and aged from 19 to 33 years. The dataset includes 159 open meshwork, 117 closed meshwork, 90 wave and 40 color fringe images.
  • The VOPTICAL_R Database distribution [Cite] contains 50 images of the preocular tear film taken over optimum illumination conditions, and acquired from healthy subjects aged from 19 to 33 years. The dataset contains images with multiple patterns, a sign of meibomian gland abnormality, including: open meshwork, closed meshwork, wave, amorphous, and color fringe.

Doane images

The VOPTICAL_GCU Database distribution contains 106 examples of real interferometric images. It includes 11 strong fringes, 25 coalescing strong fringes, 30 fine fringes, 26 coalescing fine fringes, and 14 debris images. All the images have a spatial resolution of 1280×1024 pixels, and have been acquired with the Doane interferometer and a digital PC-attached CMEX-1301 camera. This dataset has been acquired and annotated by optometrists from the Department of Life Sciences, Glasgow Caledonian University (UK). The filenames of the images indicate the predominant pattern: CO is fine fringes, FL is coalescing fine fringes, MA is strong fringes, MC is coalescing strong fringes, and AM is debris.

If you are interested in using the VOPTICAL_GCU dataset, please send an email to: bremeseiro@uniovi.es and you will receive an authentication password to access the dataset. This is for statistical purposes only, no private data or fee is required. If you use this image set for your work, please include a reference to VOPTICAL_GCU in it.

Cite

VOPTICAL_GCU, VARPA optical dataset acquired and annotated by optometrists from the Department of Life Sciences, Glasgow Caledonian University (UK), 2013. [Online] Available: http://www.varpa.org/voptical_gcu.html

Main publications

  • B. Remeseiro, N. Barreira, C. García-Resúa, M. Lira, M. J. Giráldez, E. Yebra-Pimentel, M. G. Penedo, «iDEAS: a web-based system for dry eye assessment», Computer Methods and Programs in Biomedicine, 130, 186-197, 2016.
Abstract
Background and Objectives: Dry eye disease is a public health problem, whose multifactorial etiology challenges clinicians and researchers making necessary the collaboration between different experts and centers. The evaluation of the interference patterns observed in the tear film lipid layer is a common clinical test used for dry eye diagnosis. However, it is a time-consuming task with a high degree of intra- as well as inter-observer variability, which makes the use of a computer-based analysis system highly desirable. This work introduces iDEAS (Dry Eye Assessment System), a web-based application to support dry eye diagnosis. Methods: iDEAS provides a framework for eye care experts to collaboratively work using image-based services in a distributed environment. It is composed of three main components: the web client for user interaction, the web application server for request processing, and the service module for image analysis. Specifically, this manuscript presents two automatic services: tear film classification, which classifies an image into one interference pattern; and tear film map, which illustrates the distribution of the patterns over the entire tear film. Results: iDEAS has been evaluated by specialists from different institutions to test its performance. Both services have been evaluated in terms of a set of performance metrics using the annotations of different experts. Note that the processing time of both services has been also measured for efficiency purposes. Conclusions: iDEAS is a web-based application which provides a fast, reliable environment for dry eye assessment. The system allows practitioners to share images, clinical information and automatic assessments between remote computers. Additionally, it save time for experts, diminish the inter-expert variability and can be used in both clinical and research settings.
  • B. Remeseiro, A. Mosquera, M. G. Penedo, «CASDES: A Computer-Aided System to Support Dry Eye Diagnosis Based on Tear Film Maps», IEEE Journal of Biomedical and Health Informatics, 20 (3), 936-943, 2016.
Abstract
Dry eye syndrome is recognized as a growing health problem, and one of the most frequent reasons for seeking eye care. Its etiology and management challenge clinicians and researchers alike, and several clinical tests can be used to diagnose it. One of the most frequently used tests is the evaluation of the interference patterns of the tear film lipid layer. Based on this clinical test, this paper presents CASDES, a computer-aided system to support the diagnosis of dry eye syndrome. Furthermore, CASDES is also useful to support the diagnosis of other eye diseases, such as meibomian gland dysfunction, since it provides a tear film map with highly useful information for eye practitioners. Experiments demonstrate the robustness of this novel tool, which outperforms the previous attempts to create tear film maps and provides reliable results in comparison with the clinicians’ annotations. Note that the processing time is noticeably reduced with the proposed method, which will help to promote its clinical use in the diagnosis and treatment of dry eye.
  • B. Remeseiro, V. Bolón-Canedo, D. Peteiro-Barral, A. Alonso-Betanzos, B. Guijarro-Berdiñas, A. Mosquera, M. G. Penedo, N. Sánchez-Maroño, «A Methodology for Improving Tear Film Lipid Layer Classification», IEEE Journal of Biomedical and Health Informatics, 18 (4), 1485-1493, 2014.
Abstract
Dry eye is a symptomatic disease which affects a wide range of population and has a negative impact on their daily activities. Its diagnosis can be achieved by analysing the interference patterns of the tear film lipid layer and by classifying them into one of the Guillon categories. The manual process done by experts is not only affected by subjective factors but is also very time consuming. In this research we propose a general methodology to the automatic classification of tear film lipid layer, using colour and texture information to characterise the image and feature selection methods to reduce the processing time. The adequacy of the proposed methodology was demonstrated since it achieves classification rates over 97% while maintaining robustness and provides unbiased results. Also, it can be applied in real time, and so allows important time savings for the experts.
  • B. Remeseiro, K.M. Oliver, A. Tomlinson, E. Martin, N. Barreira, A. Mosquera, «Automatic grading system for human tear films», Pattern Analysis and Applications, 18 (3), 677-694, 2015.
Abstract
Dry eye syndrome is a prevalent disease which affects a wide range of the population and has a negative impact on their daily activities, such as driving or working with computers. Its diagnosis and monitoring require a battery of tests which measure different physiological characteristics. One of these clinical tests consists in capturing the appearance of the tear film using the Doane interferometer. Once acquired, the interferometry images are classified into one of the five categories considered in this research. The variability in appearance makes the use of a computer-based analysis system highly desirable. For this reason, a general methodology for the automatic analysis and categorization of interferometry images is proposed. The development of this methodology included a deep study based on several techniques for image texture analysis, three color spaces and different machine learning algorithms. The adequacy of this methodology was demonstrated, achieving classification rates over 93%. Also, it provides unbiased results and allows important time savings for experts.
  • L. Ramos, A. Mosquera, N. Barreira, H. Pena-Verdeal, E. Yebra-Pimentel, «Break-up analysis of the tear film based on time, location, size and shape of the rupture area», 7950, 695-702, International Conference on Image Analysis and Recognition (ICIAR 13), Povoa de Varzim, Portugal, 2013.
Abstract
The Break-Up Time test (BUT) evaluates the quality and stability of the tear film. It is used for the diagnosis of the dry eye syndrome, a common disorder of the tear film, affecting a significant percentage of the population. This work describes a fully automatic methodology to compute the time in which the break-up occurs and to analyze the rupture zone. This analysis provides useful quantitative and qualitative information for the clinical practice about the location, size and shape of the break-up areas.
  • L. Ramos, A. Mosquera, N. Barreira, M. Currás, H. Pena-Verdeal, M. J. Giráldez, M. G. Penedo, «Adaptive parameter computation for the automatic measure of the Tear Break-Up Time», 243, 1370-1379, 16th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems (KES’12), San Sebastián, Setember 2012.
Abstract
Dry eye syndrome is a common disorder of the tear film, affecting a significant percentage of the population. The Break-Up Time (BUT) is a clinical test used for the diagnosis of this disease. In this work, we propose several improvements for the automatic computation of the BUT measure that solve the limitations of previous approaches. In particular, we present several procedures for the automatic computation of some parameters involved in the BUT measurement, such as the eye size, the intensity variation, or the starting point of the measurement frame sequence. We have tested our methodology on a dataset composed of 18 videos annotated by 4 different experts. The average difference between the automatic measure and the experts’ measures is on the acceptable range considering the high inter-observer variance.