Nonlinear registration for scanned retinal images: application to ocular polarimetry

Registration of polarimetric images for in vivo skin diagnostics

SIGNIFICANCE

Mueller matrix (MM) polarimetry is a promising tool for the detection of skin cancer. Polarimetric in vivo measurements often suffer from misalignment of the polarimetric images due to motion, which can lead to false results.

AIM

We aim to provide an easy-to-implement polarimetric image data registration method to ensure proper image alignment.

APPROACH

A feature-based image registration is implemented for an MM polarimeter for phantom and in vivo human skin measurements.

RESULTS

We show that the keypoint-based registration of polarimetric images is necessary for in vivo skin polarimetry to ensure reliable results. Further, we deliver an efficient semiautomated method for the registration of polarimetric images.

CONCLUSIONS

Image registration for in vivo polarimetry of human skin is required for improved diagnostics and can be efficiently enhanced with a keypoint-based approach.

Polynomial transformation model for frame-to-frame registration in an adaptive optics confocal scanning laser ophthalmoscope

The adaptive optics (AO) technique has been integrated in confocal scanning laser ophthalmoscopy (SLO) to obtain near diffraction-limited high-resolution retinal images. However, the quality of AOSLO images is decreased by various sources of noise and fixational eye movements. To improve image quality and remove distortions in AOSLO images, the multi-frame averaging method is usually utilized, which relies on an accurate image registration. The goal of image registrations is finding the optimal transformation to best align the input image sequences. However, current methods for AOSLO image registration have some obvious defects due to the limitation of transformation models. In this paper, we first established the retina motion model by using the Taylor series and polynomial expansion. Then we generated the polynomial transformation model and provided its close-form solution for consecutively frame-to-frame AOSLO retina image registration, allowing one to consider more general retinal motions such as scale changes, shearing and rotation motions, and so on. The experimental results demonstrated that higher-order polynomial transformation models are helpful to achieve more accurate registration, and the fourth-order polynomial transformation model is preferred to accomplish an efficient registration with a satisfying computational complexity. In addition, the AKAZE feature detection method was adopted and improved to achieve more accurate image registrations, and a new strategy was validated to exclude those unsuccessful registered regions to promote the robustness of image registration.

Relationships between visual field sensitivity and spectral absorption properties of the neuroretinal rim in glaucoma by multispectral imaging

PURPOSE

To investigate the relationship between neuroretinal rim (NRR) differential light absorption (DLA, a measure of spectral absorption properties) and visual field (VF) sensitivity in primary open-angle glaucoma (POAG).

METHODS

Patients diagnosed with (n = 22) or suspected of having (n = 7) POAG were imaged with a multispectral system incorporating a modified digital fundus camera, 250-W tungsten-halogen lamp, and fast-tuneable liquid crystal filter. Five images were captured sequentially within 1.0 second at wavelengths selected according to absorption properties of hemoglobin (range, 570-610 nm), and a Beer-Lambert law model was used to produce DLA maps of residual NRR from the images. Patients also underwent VF testing. Differences in NRR DLA in vertically opposing 180° and 45° sectors either side of the horizontal midline were compared with corresponding differences in VF sensitivity on both decibel and linear scales by Spearman's rank correlation.

RESULTS

The decibel VF sensitivity scale showed significant relationships between superior-inferior NRR DLA difference and sensitivity differences between corresponding VF areas in 180° NRR sectors (Spearman ρ = 0.68; P < 0.0001), superior-/inferior-temporal 45° NRR sectors (ρ = 0.57; P < 0.002), and superior-/inferior-nasal 45° NRR sectors (ρ = 0.59; P < 0.001). Using the linear VF sensitivity scale significant relationships were found for 180° NRR sectors (ρ = 0.62; P < 0.0002) and superior-inferior-nasal 45° NRR sectors (ρ = 0.53; P < 0.002). No significant difference was found between correlations using the linear or decibel VF sensitivity scales.

CONCLUSIONS

Residual NRR DLA is related to VF sensitivity in POAG. Multispectral imaging may provide clinically important information for the assessment and management of POAG.

Improving retinal image resolution with iterative weighted shift-and-add

High-resolution retinal imaging requires dilating the pupil, and therefore exposing more aberrations that blur the image. We developed an image processing technique that takes advantage of the natural movement of the eye to average out some of the high-order aberrations and to oversample the retina. This method was implemented on a long sequence of retinal images of subjects with normal vision. We were able to resolve the structures of the size of single cells in the living human retina. The improvement of resolution is independent of the acquisition method, as long as the image is not warped during scanning. Consequently, even better results can be expected by implementing this technique on higher-resolution images.

Scanning ophthalmoscope retinal image registration using one-dimensional deformation fields

We present a new, robust and automated method for registering sequences of images acquired from scanning ophthalmoscopes. The method uses a multi-scale B-spline representation of the deformation field to map images to each other and an hierarchical optimization method. We applied the method to video sequences acquired from different parts of the retina. In all cases, the registration was successful, even in the presence of large distortions from microsaccades, and the resulting deformation fields describe the fixational motion of the eye. The registration reveals examples of dynamic photoreceptor behaviour in the sequences.

Geodesic distance on non-singular coherency matrix space in polarization optics

We define a geodesic distance associated with the polarization space of non-singular coherency matrices. Its introduction on HPD(2) (the manifold of Hermitian positive definite matrices of dimension 2) can be directly related to the Jones calculus. The expression of distance and related notion of mean value in this particular metric space are also presented. We investigate the properties of this geodesic distance and the classical Euclidean one and their appropriateness for interpixel comparisons in a context of imaging polarimetry. Finally, results are presented for a geodesic version of the classical K-means clustering algorithm with simulated data and real data. The results demonstrate the advantages of the geodesic approach.