Automatic segmentation of the foveal avascular zone in ophthalmological OCT-A images

Abstract

Angiography by Optical Coherence Tomography (OCT-A) is a non-invasive retinal imaging modality of recent appearance that allows the visualization of the vascular structure at predefined depths based on the detection of the blood movement through the retinal vasculature. In this way, OCT-A images constitute a suitable scenario to analyze the retinal vascular properties of regions of interest as is the case of the macular area, measuring the characteristics of the foveal vascular and avascular zones. Extracted parameters of this region can be used as prognostic factors that determine if the patient suffers from certain pathologies (such as diabetic retinopathy or retinal vein occlusion, among others), indicating the associated pathological degree. The manual extraction of these biomedical parameters is a long, tedious and subjective process, introducing a significant intra and inter-expert variability, which penalizes the utility of the measurements. In addition, the absence of tools that automatically facilitate these calculations encourages the creation of computer-aided diagnosis frameworks that ease the doctor's work, increasing their productivity and making viable the use of this type of vascular biomarkers. In this work we propose a fully automatic system that identifies and precisely segments the region of the foveal avascular zone (FAZ) using a novel ophthalmological image modality as is OCT-A. The system combines different image processing techniques to firstly identify the region where the FAZ is contained and, secondly, proceed with the extraction of its precise contour. The system was validated using a representative set of 213 healthy and diabetic OCT-A images, providing accurate results with the best correlation with the manual measurements of two experts clinician of 0.93 as well as a Jaccard's index of 0.82 of the best experimental case in the experiments with healthy OCT-A images. The method also provided satisfactory results in diabetic OCT-A images, with a best correlation coefficient with the manual labeling of an expert clinician of 0.93 and a Jaccard's index of 0.83. This tool provides an accurate FAZ measurement with the desired objectivity and reproducibility, being very useful for the analysis of relevant vascular diseases through the study of the retinal micro-circulation.

Fig 1. Examples of OCT-A images representing all the configurations that were used in this work.

1^st row, images of 3x3 millimeters. 2^nd row, images of 6x6 millimeters. (a) & (c) Superficial OCT-A images. (b) & (d), Deep OCT-A images.

Fig 2. Main steps of the proposed methodology.

Fig 3. Application of the preprocessing step.

(a) Original image. (b) Image result after applying the-top hat operator.

Fig 4. Vascularity edge identification using the Canny edge detector.

(a) Original OCT-A image (after the top-hat preprocessing step). (b) Results of the vascular edge identification.

Fig 5. Morphological closure and inversion of intensities followed by a removal of small elements.

(a) Image with the vascular edge identification. (b) Result after applying a morphological closure. (c) Result after applying an inversion of intensity and an opening.

Fig 6. Example of error in the capture process.

(a) Original image. (b) Initial set of identified FAZ candidates. (c) Final set of FAZ candidates after FP removal.

Fig 7. Removal process of FAZ FP candidates.

(a) Initial set of identified FAZ candidates. (b) Final set of FAZ candidates.

Fig 8. Application of the precise final FAZ segmentation.

(a) & (c) Preliminary FAZ extractions. (b) & (d) Final segmentation results.

Fig 9. Error cases on the FAZ localization process.

Fig 10. Comparative examples of the experts (green and red) and the automatic computational (blue) segmentations as well as the corresponding area size measurements.

Fig 11. Comparative examples with goods and bad results in the Jaccard’s index in the four subgroups (superficial and deep in 3 × 3 and 6 × 6 sizes).

Fig 12. Examples of representative FAZ regions from the defined levels of circularity in the diabetic OCT-A dataset.

(a) High level of circularity, (b) medium level of circularity and (c) low level of circularity.

Fig 13. Comparative examples of the experts (green and red) and the automatic computational (blue) FAZ measurements in superficial 3 millimeters images.

Fig 14. Comparative examples of the experts (green and red) and the automatic computational (blue) FAZ measurements in superficial 6 millimeters images.

Fig 15. Comparative examples of the experts (green and red) and the automatic computational (blue) FAZ measurements in deep 3 millimeters images.

Fig 16. Comparative examples of the experts (green and red) and the automatic computational (blue) FAZ measurements in deep 6 millimeters images.