Geometric Correspondence-Based Multimodal Learning for Ophthalmic Image Analysis.

Color fundus photography (CFP) and Optical coherence tomography (OCT) images are two of the most widely used modalities in the clinical diagnosis and management of retinal diseases. Despite the widespread use of multimodal imaging in clinical practice, few methods for automated diagnosis of eye diseases utilize correlated and complementary information from multiple modalities effectively. This paper explores how to leverage the information from CFP and OCT images to improve the automated diagnosis of retinal diseases. We propose a novel multimodal learning method, named geometric correspondence-based multimodal learning network (GeCoM-Net), to achieve the fusion of CFP and OCT images. Specifically, inspired by clinical observations, we consider the geometric correspondence between the OCT slice and the CFP region to learn the correlated features of the two modalities for robust fusion. Furthermore, we design a new feature selection strategy to extract discriminative OCT representations by automatically selecting the important feature maps from OCT slices. Unlike the existing multimodal learning methods, GeCoM-Net is the first method that formulates the geometric relationships between the OCT slice and the corresponding region of the CFP image explicitly for CFP and OCT fusion. Experiments have been conducted on a large-scale private dataset and a publicly available dataset to evaluate the effectiveness of GeCoM-Net for diagnosing diabetic macular edema (DME), impaired visual acuity (VA) and glaucoma. The empirical results show that our method outperforms the current state-of-the-art multimodal learning methods by improving the AUROC score 0.4%, 1.9% and 2.9% for DME, VA and glaucoma detection, respectively.

Are macular drusen in midlife a marker of accelerated biological ageing?

CLINICAL RELEVANCE: Macular drusen are associated with age-related maculopathy but are not an ocular manifestation or biomarker of systemic ageing. BACKGROUND: Macular drusen are the first sign of age-related maculopathy, an eye disease for which age is the strongest risk factor. The aim of this cohort study was to investigate whether macular drusen in midlife - a sign of the earliest stages of age-related macular degeneration (AMD) - are associated with accelerated biological ageing more generally. METHODS: Members of the long-running Dunedin Multidisciplinary Health and Development Study (hereafter the Dunedin Study, n = 1037) underwent retinal photography at their most recent assessment at the age of 45 years. Images were graded for the presence of AMD using a simplified scale from the Age-Related Eye Disease Study (AREDS). Accelerated ageing was assessed by (i) a measure of Pace of Ageing defined from a combination of clinical and serum biomarkers obtained at ages 26, 32, 38, and 45 years and (ii) Facial Ageing, defined from photographs obtained at age 38 and 45 years. RESULTS: Of the 938 participants who participated at the age 45 assessments, 834 had gradable retinal photographs, and of these 165 (19.8%) had macular drusen. There was no significant difference in Pace of Ageing (p = .743) or Facial Ageing (p = .945) among participants with and without macular drusen. CONCLUSIONS: In this representative general population sample, macular drusen in midlife were not associated with accelerated ageing. Future studies tracking longitudinal changes in drusen number and severity at older ages may reveal whether drusen are a biomarker of accelerated ageing.

Association of macular and choroidal perfusion with long-term visual outcomes after macula-off rhegmatogenous retinal detachment.

BACKGROUND/AIMS: To examine the relationship between macular perfusion, as assessed using optical coherence tomography angiography (OCTA), and long-term visual outcome after surgical repair of macula-off rhegmatogenous retinal detachment (RRD). METHODS: A prospective study of 29 patients who had undergone successful surgical repair of macula-off RRD. OCTA imaging was performed at month 3 and repeated at months 6 and 12 after surgery. Associations between OCTA parameters including, foveal avascular zone (FAZ) area, vessel density (VD) in the superficial capillary plexus (SCP) and deep capillary plexus (DCP), choriocapillaris flow deficit features and logMAR best-corrected visual acuity (VA) were assessed using a random intercept hybrid linear mixed model. RESULTS: Over the 1-year follow-up, VA improved (0.025 logMAR/ month, 95% CI 0.015 to 0.035) and FAZ area decreased (-0.020 mm2/month, 95% CI -0.032 to -0.007). Better VA after surgery was significantly associated with denser superficial VD (ß=0.079, 95% CI 0.026 to 0.131), lower number of choriocapillaris flow deficits (ß=-0.087, 95% CI -0.154 to -0.021) and larger average size of choriocapillaris flow deficits (ß=0.085, 95% CI 0.022 to 0.147), after adjusting for baseline VA, types of surgery and other factors. CONCLUSIONS: OCTA measures of vascular perfusion in the macula may provide new pathophysiological insights and prognostic information related to macula-off RRD.