Article

A novel registration method for retinal images based on local features.

Institute of Automation, Chinese Academy of Science, Beijing, China.
Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 02/2008; 2008:2242-5. DOI: 10.1109/IEMBS.2008.4649642
Source: IEEE Xplore

ABSTRACT Sometimes it is very hard to automatically detect the bifurcations of vascular network in retinal images so that the general feature based registration methods will fail to register two images. In order to solve this problem, we developed a novel local feature based retinal image registration method. We first detect the corner points instead of bifurcations since corner points are sufficient and uniformly distributed in the overlaps. Second, a novel highly distinctive local feature is extracted around each corner point. These local features are invariant to rotation and contrast, and partially invariant to scaling. Third, a bilateral matching technique is applied to identify the corresponding features between two images. Finally a second order polynomial transformation is used to register two images. Experimental results show that our method is very robust and compute efficient to register retinal images even of very low quality.

0 Bookmarks
 · 
105 Views
  • [Show abstract] [Hide abstract]
    ABSTRACT: Purpose: To evaluate geographic atrophy (GA) progression in eyes with dry age-related macular degeneration (AMD) and to determine factors related to GA expansion, notably reticular pseudodrusen (RPD), also known as subretinal drusenoid deposits (SDD) or reticular macular disease (RMD). Methods: This was a retrospective cohort study of patients with dry AMD who were diagnosed with GA in at least 1 eye and were imaged with sequential fundus autofluorescence (FAF) and/or near infrared reflectance (NIR-R) imaging. Images were analyzed for the presence of GA within the macular region. GA progression was measured in the fields of a modified Wisconsin grid and spatially correlated with RPD. Factors also evaluated for association with GA progression included initial GA size and pattern. Results: The study sample included 126 eyes of 92 patients, with an average follow-up of 20.4 months (SD=11.7). At baseline, 93.6% of eyes had RPD, and the average GA area was 2.8 mm2 (SD=2.9). The average GA progression rate was 0.8 mm2/year (SD=0.6), with a statistically significant difference between the unilobular and multilobular phenotype groups (0.3 mm2/year vs. 0.9 mm2/year, p=0.02). Patients in the lower 50th percentile of initial GA area had a lower progression rate than patients in the upper 50th percentile (0.6 mm2/year vs. 1.1 mm2/year, p<0.001). GA progression was more frequent in fields with RPD than in those without RPD (74.2% vs. 41.7%, p<0.001). Conclusions: The high correlation between the presence of RPD (also known as SDD or RMD) and the presence of GA and the expansion of GA into areas with these lesions suggest that they are an early manifestation of the process leading to GA.
    Investigative ophthalmology & visual science 10/2013; · 3.43 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Computer-aided intraocular surgery requires precise, real-time knowledge of the vasculature during retinal procedures such as laser photocoagulation or vessel cannulation. Because vitreoretinal surgeons manipulate retinal structures on the back of the eye through ports in the sclera, voluntary and involuntary tool motion rotates the eye in the socket and causes movement to the microscope view of the retina. The dynamic nature of the surgical workspace during intraocular surgery makes mapping, tracking, and localizing vasculature in real time a challenge. We present an approach that both maps and localizes retinal vessels by temporally fusing and registering individual-frame vessel detections. On video of porcine and human retina, we demonstrate real-time performance, rapid convergence, and robustness to variable illumination and tool occlusion.
    IEEE Int Conf Robot Autom 01/2013;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Glaucoma is the leading cause of preventable blindness in the western world. Investigation of high-resolution retinal nerve fiber layer (RNFL) images in patients may lead to new indicators of its onset. Adaptive optics (AO) can provide diffraction-limited images of the retina, providing new opportunities for earlier detection of neuroretinal pathologies. However, precise processing is required to correct for three effects in sequences of AO-assisted, flood-illumination images: uneven illumination, residual image motion and image rotation. This processing can be challenging for images of the RNFL due to their low contrast and lack of clearly noticeable features. Here we develop specific processing techniques and show that their application leads to improved image quality on the nerve fiber bundles. This in turn improves the reliability of measures of fiber texture such as the correlation of Gray-Level Co-occurrence Matrix (GLCM).
    Biomedical Optics Express 06/2014; 5(6):1941-51. · 3.50 Impact Factor

Preview (2 Sources)

Download
1 Download
Available from