Glaucoma, a leading cause of blindness, is a multifaceted disease with several patho-physiological features manifesting in single fundus images (e.g., optic nerve cupping) as well as fundus videos (e.g., vascular pulsatility index). Current convolutional neural networks (CNNs) developed to detect glaucoma are all based on spatial features embedded in an image. We developed a combined CNN and recurrent neural network (RNN) that not only extracts the spatial features in a fundus image but also the temporal features embedded in a fundus video (i.e., sequential images). A total of 1810 fundus images and 295 fundus videos were used to train a CNN and a combined CNN and Long Short-Term Memory RNN. The combined CNN/RNN model reached an average F-measure of 96.2% in separating glaucoma from healthy eyes. In contrast, the base CNN model reached an average F-measure of only 79.2%. This proof-of-concept study demonstrates that extracting spatial and temporal features from fundus videos using a combined CNN and RNN, can markedly enhance the accuracy of glaucoma detection.

Retinal imaging may serve as an alternative approach to monitor brain pathology in Alzheimer’s disease (AD). In this study, we investigated the association between retinal vascular and structural changes and cerebral amyloid-β (Aβ) plaque load in an elderly cohort. We studied a total of 101 participants, including 73 elderly subjects (79 ± 5 years, 22 male) with no clinical diagnosis of AD but reporting some subjective memory change and an additional 28 subjects (70 ± 9 years, 16 male) with clinically established AD. Following a complete dilated ocular examination, the amplitude of retinal vascular pulsations and dynamic response, retinal nerve fibre layer thickness and retinal ganglion cell layer (RGCL) thickness were determined in all patients. Systemic blood pressure and carotid-to-femoral pulse wave velocity were measured. The elderly cohort also underwent magnetic resonance imaging and 18F-florbetaben (FBB)-positron emission tomographic amyloid imaging to measure neocortical Aβ standardised uptake value ratio (SUVR), and this was used to characterise a ‘preclinical’ group (SUVR >1.4). The mean FBB neocortical SUVR was 1.35 ± 0.3. The amplitude of retinal venous pulsations correlated negatively with the neocortical Aβ scores (p < 0.001), whereas the amplitude of retinal arterial pulsations correlated positively with neocortical Aβ scores (p < 0.01). RGCL thickness was significantly lower in the clinical AD group (p < 0.05). The correlation between retinal vascular changes and Aβ plaque load supports the possibility of a vascular component to AD. Dynamic retinal vascular parameters may provide an additional inexpensive tool to aid in the preclinical assessment of AD.