Multi-modal automatic montaging of adaptive optics retinal images.

We present a fully automated adaptive optics (AO) retinal image montaging algorithm using classic scale invariant feature transform with random sample consensus for outlier removal. Our approach is capable of using information from multiple AO modalities (confocal, split detection, and dark field) and can accurately detect discontinuities in the montage. The algorithm output is compared to manual montaging by evaluating the similarity of the overlapping regions after montaging, and calculating the detection rate of discontinuities in the montage. Our results show that the proposed algorithm has high alignment accuracy and a discontinuity detection rate that is comparable (and often superior) to manual montaging. In addition, we analyze and show the benefits of using multiple modalities in the montaging process. We provide the algorithm presented in this paper as open-source and freely available to download.

[1]  Hao Li,et al.  Automatic montage of retinal images in adaptive optics confocal scanning laser ophthalmoscope , 2012 .

[2]  Jean-Michel Morel,et al.  ASIFT: A New Framework for Fully Affine Invariant Image Comparison , 2009, SIAM J. Imaging Sci..

[3]  Juan Carlos Niebles,et al.  Unsupervised Learning of Human Action Categories Using Spatial-Temporal Words , 2008, International Journal of Computer Vision.

[4]  T. Hebert,et al.  Adaptive optics scanning laser ophthalmoscopy. , 2002, Optics express.

[5]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[6]  Brian B. Avants,et al.  Symmetric diffeomorphic image registration with cross-correlation: Evaluating automated labeling of elderly and neurodegenerative brain , 2008, Medical Image Anal..

[7]  Max A. Viergever,et al.  Mutual-information-based registration of medical images: a survey , 2003, IEEE Transactions on Medical Imaging.

[8]  Joydeep Ghosh,et al.  Cluster Ensembles --- A Knowledge Reuse Framework for Combining Multiple Partitions , 2002, J. Mach. Learn. Res..

[9]  A. Dubra,et al.  In vivo dark-field imaging of the retinal pigment epithelium cell mosaic. , 2013, Biomedical optics express.

[10]  A. Dubra,et al.  Reflective afocal broadband adaptive optics scanning ophthalmoscope , 2011, Biomedical optics express.

[11]  Toco Y P Chui,et al.  The use of forward scatter to improve retinal vascular imaging with an adaptive optics scanning laser ophthalmoscope , 2012, Biomedical optics express.

[12]  D R Williams,et al.  Supernormal vision and high-resolution retinal imaging through adaptive optics. , 1997, Journal of the Optical Society of America. A, Optics, image science, and vision.

[13]  David Williams,et al.  Noninvasive imaging of the human rod photoreceptor mosaic using a confocal adaptive optics scanning ophthalmoscope , 2011, Biomedical optics express.

[14]  Christopher S. Langlo,et al.  In vivo imaging of human cone photoreceptor inner segments. , 2014, Investigative ophthalmology & visual science.

[15]  Luc Van Gool,et al.  Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..

[16]  Jessica I W Morgan,et al.  The fundus photo has met its match: optical coherence tomography and adaptive optics ophthalmoscopy are here to stay , 2016, Ophthalmic & physiological optics : the journal of the British College of Ophthalmic Opticians.

[17]  Nikos Paragios,et al.  Deformable Medical Image Registration: A Survey , 2013, IEEE Transactions on Medical Imaging.

[18]  P. Lennie,et al.  Packing arrangement of the three cone classes in primate retina , 2001, Vision Research.

[19]  Jessica I. W. Morgan,et al.  In vivo autofluorescence imaging of the human and macaque retinal pigment epithelial cell mosaic. , 2009, Investigative ophthalmology & visual science.