Page 16 - Computational Retinal Image Analysis
P. 16

CHAPTER


                  Clinical motivation and the

                  needs for RIA in healthcare                                2







                                                                                    b
                                                                     a
                                                          Ryo Kawasaki , Jakob Grauslund
                   a Department of Vision Informatics, Osaka University Graduate School of Medicine, Suita, Japan,
                  b Department of Ophthalmology and Department of Clinical Research, Odense University Hospital
                                               and University of Southern Denmark, Odense, Denmark



                  1  Introduction
                  The retina provides a unique window to observe blood vessels and neural tissue
                  in vivo. The intervention of the ophthalmoscope by Hermann von Helmholtz in 1851
                  has uncovered the retina behind the pupil of the eye [1]. Observing the retina and the
                  optic disc has formed the foundation of modern ophthalmology by describing ocular
                  diseases in detail.
                     Ophthalmologists diagnose retinal diseases by identifying specific signs on the
                  retina. In the clinical setting, the primarily method to examine the retina is through
                  direct or indirect ophthalmoscopy. Expanding use of retinal photographs contributed
                  to retinal image analysis tremendously. Not long since the invention of photography,
                  Jackman and Webster reported the first fundus photography in 1886 [2]. Carl Zeiss
                  produced the first commercial fundus camera in 1926. Since then, fundus photography
                  has been used as the main imaging modality in clinical settings. In the 1970s, the non-
                  mydriatic fundus camera was introduced to the market and it contributed to expanding
                  the field of fundus photography use to health screening programs or epidemiological
                  studies. In the early days, retinal image analysis was done by literally applying scales
                  or a Vernier caliper on the films or projected images on a large screen (e.g. measur-
                  ing vessel caliber, optic disc diameter, and standard circles for signs of age-related
                  macular degeneration). At that time, as one can imagine, retinal image analyses were
                  limited because of its workload and so mainly used for research purposes.
                     Landmark expansion of retinal imaging was achieved by a transition from film to
                  digital image. Digital fundus camera was introduced by Eastman Kodak in 1975. It
                  took several decades to fully shift from analog film to digital imaging by persuading
                  clinicians to accept color replication and high resolution to capture subtle changes on
                  the retinal images (and maybe cost of the equipment). Digital fundus photography is
                  now the standard imaging modality in both ophthalmic clinics and screening facili-
                  ties. Since the inclusion of a digital camera along with matching computing power



                  Computational Retinal Image Analysis. https://doi.org/10.1016/B978-0-08-102816-2.00002-2  5
                  © 2019 Elsevier Ltd. All rights reserved.
   11   12   13   14   15   16   17   18   19   20   21